Initial commit

.dockerignore

@@ -0,0 +1,75 @@
+# Python cache files
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+env/
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# Virtual environments
+.venv/
+venv/
+ENV/
+env/
+
+# IDE files
+.vscode/
+.idea/
+*.swp
+*.swo
+*~
+
+# OS files
+.DS_Store
+Thumbs.db
+
+# Git files
+.git/
+.gitignore
+
+# Project specific
+data/chroma_db/
+data/*.log
+uploads/*
+!uploads/.gitkeep
+
+# Environment files
+.env
+
+# Documentation
+*.md
+!README.md
+
+# Lock files
+uv.lock
+poetry.lock
+Pipfile.lock
+
+# Testing
+.pytest_cache/
+.coverage
+htmlcov/
+
+# Jupyter notebooks
+*.ipynb
+.ipynb_checkpoints/
+
+# Local development
+local/
+tmp/
+temp/

.env.example

@@ -0,0 +1,92 @@
+# =============================================================================
+# LEGA.AI CONFIGURATION TEMPLATE
+# =============================================================================
+# Copy this file to .env and update the values below
+# This is a template file - safe to commit to version control
+
+# =============================================================================
+# GOOGLE AI API CONFIGURATION (REQUIRED)
+# =============================================================================
+# Get your API key from: https://aistudio.google.com/
+GOOGLE_API_KEY=your-google-api-key-here
+
+# =============================================================================
+# APPLICATION SETTINGS
+# =============================================================================
+DEBUG=True
+LOG_LEVEL=INFO
+
+# Streamlit server configuration
+STREAMLIT_SERVER_PORT=8501
+STREAMLIT_SERVER_ADDRESS=localhost
+
+# File upload settings
+MAX_FILE_SIZE_MB=10
+SUPPORTED_FILE_TYPES=pdf,docx,txt
+
+# =============================================================================
+# AI MODEL SETTINGS
+# =============================================================================
+TEMPERATURE=0.2
+MAX_TOKENS=2048
+EMBEDDING_MODEL=models/text-embedding-004
+
+# =============================================================================
+# VECTOR STORE CONFIGURATION
+# =============================================================================
+CHROMA_PERSIST_DIRECTORY=./data/chroma_db
+
+# =============================================================================
+# STORAGE CONFIGURATION
+# =============================================================================
+UPLOAD_DIR=./uploads
+DATA_DIR=./data
+LOG_FILE=./data/app.log
+
+# =============================================================================
+# SECURITY SETTINGS
+# =============================================================================
+SECRET_KEY=your-secret-key-here
+SESSION_TIMEOUT_MINUTES=60
+
+# =============================================================================
+# DATABASE CONFIGURATION
+# =============================================================================
+DATABASE_URL=sqlite:///./data/lega.db
+
+# =============================================================================
+# PERFORMANCE SETTINGS
+# =============================================================================
+MAX_CONCURRENT_UPLOADS=5
+DOCUMENT_PROCESSING_TIMEOUT=300
+ENABLE_CACHE=True
+CACHE_TTL_SECONDS=3600
+
+# =============================================================================
+# FEATURE FLAGS
+# =============================================================================
+ENABLE_DOCUMENT_LIBRARY=True
+ENABLE_QA_ASSISTANT=True
+ENABLE_MARKET_COMPARISON=True
+ENABLE_TIMELINE_TRACKER=True
+ENABLE_EXPORT_FEATURES=True
+
+# =============================================================================
+# REGIONAL SETTINGS
+# =============================================================================
+DEFAULT_REGION=India
+DEFAULT_CURRENCY=INR
+TIMEZONE=Asia/Kolkata
+
+# =============================================================================
+# ADVANCED AI SETTINGS
+# =============================================================================
+RISK_SENSITIVITY=3
+SIMPLIFICATION_LEVEL=intermediate
+MAX_RISK_FACTORS=10
+
+# =============================================================================
+# API RATE LIMITING
+# =============================================================================
+API_REQUESTS_PER_MINUTE=60
+API_REQUESTS_PER_DAY=1000

.env.hf

@@ -0,0 +1,51 @@
+# =============================================================================
+# HUGGING FACE SPACES CONFIGURATION
+# =============================================================================
+# This file contains default configuration for Hugging Face Spaces deployment
+# Set these environment variables in your Hugging Face Space settings
+
+# =============================================================================
+# REQUIRED: GOOGLE AI API CONFIGURATION
+# =============================================================================
+# Get your API key from: https://aistudio.google.com/
+# Set this in Hugging Face Spaces under Settings -> Variables
+GOOGLE_API_KEY=
+
+# =============================================================================
+# HUGGING FACE SPACES SETTINGS (Automatically configured)
+# =============================================================================
+# These are set automatically by the startup script
+PORT=7860
+STREAMLIT_SERVER_PORT=7860
+STREAMLIT_SERVER_ADDRESS=0.0.0.0
+
+# =============================================================================
+# APPLICATION SETTINGS FOR PRODUCTION
+# =============================================================================
+DEBUG=False
+LOG_LEVEL=INFO
+
+# File upload settings (conservative for cloud deployment)
+MAX_FILE_SIZE_MB=5
+SUPPORTED_FILE_TYPES=pdf,docx,txt
+
+# =============================================================================
+# AI MODEL SETTINGS (Optimized for Hugging Face)
+# =============================================================================
+TEMPERATURE=0.2
+MAX_TOKENS=2048
+EMBEDDING_MODEL=models/text-embedding-004
+
+# =============================================================================
+# STORAGE CONFIGURATION (Cloud optimized)
+# =============================================================================
+CHROMA_PERSIST_DIRECTORY=./data/chroma_db
+UPLOAD_DIR=./uploads
+DATA_DIR=./data
+LOG_FILE=./data/app.log
+
+# =============================================================================
+# SECURITY SETTINGS FOR CLOUD DEPLOYMENT
+# =============================================================================
+SECRET_KEY=huggingface-lega-ai-deployment
+SESSION_TIMEOUT_MINUTES=30

.gitignore

@@ -0,0 +1,32 @@
+# Python-generated files
+__pycache__/
+*.py[oc]
+build/
+dist/
+wheels/
+*.egg-info
+
+# Virtual environments
+.venv
+
+# Environment variables (contains secrets)
+.env
+
+# Local data and uploads
+data/
+uploads/
+*.log
+
+# Temporary files
+temp/
+tmp/
+.tmp
+
+# IDE files
+.vscode/
+.idea/
+*.swp
+*.swo
+
+# OS files
+Thumbs.db

.python-version

@@ -0,0 +1 @@
+3.13

.streamlit/config.toml

@@ -0,0 +1,17 @@
+[server]
+port = 7860
+address = "0.0.0.0"
+headless = true
+fileWatcherType = "none"
+enableCORS = false
+enableXsrfProtection = false
+
+[browser]
+gatherUsageStats = false
+
+[global]
+dataFrameSerialization = "legacy"
+
+[client]
+caching = false
+displayEnabled = false

Dockerfile

@@ -0,0 +1,47 @@
+# Use Python 3.11 as Hugging Face Spaces supports it well
+FROM python:3.11-slim
+
+# Set working directory
+WORKDIR /app
+
+# Set environment variables for Python
+ENV PYTHONUNBUFFERED=1 \
+    PYTHONDONTWRITEBYTECODE=1 \
+    PIP_NO_CACHE_DIR=1 \
+    PIP_DISABLE_PIP_VERSION_CHECK=1
+
+# Install system dependencies
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    git \
+    && rm -rf /var/lib/apt/lists/*
+
+# Copy requirements first for better layer caching
+COPY requirements.txt .
+
+# Install Python dependencies
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy the application code
+COPY . .
+
+# Create necessary directories
+RUN mkdir -p data/chroma_db uploads data .streamlit && \
+    touch data/app.log && \
+    chmod -R 777 data uploads .streamlit
+
+# Expose the port that Streamlit runs on
+EXPOSE 7860
+
+# Set the default port for Hugging Face Spaces
+ENV STREAMLIT_SERVER_PORT=7860
+ENV STREAMLIT_SERVER_ADDRESS=0.0.0.0
+ENV STREAMLIT_CONFIG_DIR=/app/.streamlit
+ENV XDG_CONFIG_HOME=/app
+
+# Health check
+HEALTHCHECK CMD curl --fail http://localhost:7860/_stcore/health
+
+# Command to run the application
+CMD ["streamlit", "run", "main.py", "--server.port=7860", "--server.address=0.0.0.0", "--server.headless=true", "--server.fileWatcherType=none"]

HUGGINGFACE_DEPLOYMENT.md

@@ -0,0 +1,83 @@
+# Hugging Face Spaces Deployment Guide for Lega.AI
+
+## 🚀 Quick Deployment to Hugging Face Spaces
+
+### Step 1: Create a New Space
+
+1. Go to [Hugging Face Spaces](https://huggingface.co/spaces)
+2. Click "Create new Space"
+3. Choose:
+   - **Space name**: `lega-ai` (or your preferred name)
+   - **License**: `MIT`
+   - **SDK**: `Docker`
+   - **Hardware**: `CPU basic` (sufficient for this app)
+
+### Step 2: Upload the Code
+
+1. Clone or download this repository
+2. Upload all files to your Hugging Face Space repository
+3. Ensure the `README.md` has the correct frontmatter:
+   ```yaml
+   ---
+   title: Lega.AI
+   emoji: ⚖️
+   colorFrom: pink
+   colorTo: indigo
+   sdk: docker
+   pinned: false
+   ---
+   ```
+
+### Step 3: Configure Environment Variables
+
+1. In your Space, go to **Settings** → **Variables**
+2. Add the required environment variable:
+   - **Name**: `GOOGLE_API_KEY`
+   - **Value**: Your Google AI API key from [Google AI Studio](https://aistudio.google.com/)
+
+### Step 4: Deploy
+
+1. Commit and push your changes to the Space repository
+2. Hugging Face will automatically build and deploy your Docker container
+3. Wait for the build to complete (usually 5-10 minutes)
+4. Your app will be available at `https://huggingface.co/spaces/[username]/[space-name]`
+
+## 🔧 Customization Options
+
+### Environment Variables You Can Set:
+
+- `GOOGLE_API_KEY` (required)
+- `MAX_FILE_SIZE_MB` (default: 5)
+- `TEMPERATURE` (default: 0.2)
+- `LOG_LEVEL` (default: INFO)
+
+### Hardware Requirements:
+
+- **CPU Basic**: Sufficient for most use cases
+- **CPU Upgrade**: Recommended for heavy usage
+- **GPU**: Not required for this application
+
+## 📋 Troubleshooting
+
+### Common Issues:
+
+1. **Build fails**: Check that all files are uploaded correctly
+2. **API errors**: Ensure `GOOGLE_API_KEY` is set correctly
+3. **Timeout**: Consider upgrading to CPU Upgrade hardware
+
+### Logs:
+
+- Check the Space logs in the Hugging Face interface
+- Look for startup messages and error information
+
+## 🔒 Security Considerations
+
+- Never commit your API key to the repository
+- Use Hugging Face Spaces environment variables for sensitive data
+- The application runs in a sandboxed environment on Hugging Face
+
+## 📊 Usage Limits
+
+- Hugging Face Spaces has usage limits for free tiers
+- Consider upgrading for production use
+- Monitor usage in your Hugging Face account dashboard

README.md

@@ -0,0 +1,484 @@
+---
+title: Lega.AI
+emoji: ⚖️
+colorFrom: pink
+colorTo: indigo
+sdk: docker
+pinned: false
+---
+
+# Lega.AI
+
+AI-powered legal document analysis and simplification platform that makes complex legal documents accessible to everyone.
+
+![Python](https://img.shields.io/badge/Python-3.13+-blue.svg)
+![Streamlit](https://img.shields.io/badge/Streamlit-1.49+-red.svg)
+![LangChain](https://img.shields.io/badge/LangChain-0.3+-green.svg)
+![License](https://img.shields.io/badge/License-MIT-yellow.svg)
+
+## 📋 Table of Contents
+
+- [🚀 Features](#-features)
+- [🛠️ Tech Stack](#️-tech-stack)
+- [📋 Prerequisites](#-prerequisites)
+- [🚀 Quick Start](#-quick-start)
+- [🐳 Docker Deployment](#-docker-deployment)
+- [📁 Project Structure](#-project-structure)
+- [🎯 Usage Guide](#-usage-guide)
+- [📄 Sample Documents](#-sample-documents)
+- [🚨 Document Types Supported](#-document-types-supported)
+- [⚡ Key Features Deep Dive](#-key-features-deep-dive)
+- [🔧 Configuration Options](#-configuration-options)
+- [🔒 Privacy & Security](#-privacy--security)
+- [🤝 Contributing](#-contributing)
+- [🆘 Support](#-support)
+- [🎯 Roadmap](#-roadmap)
+
+## 🚀 Features
+
+- **🔍 Advanced Document Analysis**: Upload PDF/DOCX/TXT files and get comprehensive AI-powered analysis using Google's Gemini
+- **📝 Plain Language Translation**: Convert complex legal jargon into clear, understandable language with context-aware explanations
+- **⚠️ Intelligent Risk Assessment**: Multi-dimensional risk scoring with color-coded severity levels and detailed explanations
+- **💬 Interactive Q&A Assistant**: Ask specific questions about your documents and get instant, context-aware AI responses
+- **🎯 Smart Clause Highlighting**: Visual highlighting of risky clauses with interactive tooltips and improvement suggestions
+- **📊 Vector-Powered Similarity Search**: Find similar clauses across documents using Chroma vector database
+- **📚 Persistent Document Library**: Organize, search, and manage all analyzed documents with metadata
+- **⚠️ Risk Visualization**: Interactive charts and gauges showing risk distribution and severity
+- **🗓️ Key Information Extraction**: Automatically identify important dates, deadlines, and financial terms
+- **💾 Local Data Persistence**: Secure local storage of analysis results and vector embeddings
+- **🎨 Modern UI/UX**: Responsive Streamlit interface with custom CSS and intuitive navigation
+
+## 🛠️ Tech Stack
+
+- **Frontend**: Streamlit with multi-page navigation and custom CSS styling
+- **AI/ML**: LangChain + Google Generative AI (Gemini Pro)
+- **Embeddings**: Google Generative AI Embeddings (models/text-embedding-004)
+- **Vector Store**: Chroma for document similarity search and persistence
+- **Document Processing**: PyPDF for PDF extraction, python-docx for Word documents
+- **Package Management**: UV (modern Python package manager)
+- **Configuration**: Python-dotenv for environment management
+- **Visualization**: Plotly for interactive charts and analytics
+- **UI Components**: Streamlit-option-menu for enhanced navigation
+
+## 📋 Prerequisites
+
+- Python 3.13+ (required for latest features and performance)
+- Google AI API key (get from [Google AI Studio](https://aistudio.google.com/))
+- UV package manager (recommended for fast, reliable dependency management)
+
+## 🚀 Quick Start
+
+### 1. **Clone and navigate to the project**:
+
+```bash
+git clone <repository-url>
+cd Lega.AI
+```
+
+### 2. **Install UV (if not already installed)**:
+
+```bash
+# On macOS/Linux
+curl -LsSf https://astral.sh/uv/install.sh | sh
+
+# On Windows (PowerShell)
+powershell -c "irm https://astral.sh/uv/install.ps1 | iex"
+
+# Or using pip
+pip install uv
+```
+
+### 3. **Set up environment and install dependencies**:
+
+```bash
+# Create and activate virtual environment with dependencies
+uv sync
+
+# Or if you prefer traditional approach:
+# uv venv
+# source .venv/bin/activate  # On Windows: .venv\Scripts\activate
+# uv pip install -r pyproject.toml
+```
+
+### 4. **Configure environment**:
+
+```bash
+# Copy the template file
+cp .env.example .env
+
+# Edit .env file and update the following required settings:
+```
+
+**Required Configuration:**
+
+```env
+# Get your API key from: https://aistudio.google.com/
+GOOGLE_API_KEY=your-google-api-key-here
+```
+
+**Optional Configuration (with sensible defaults):**
+
+```env
+# Application Settings
+DEBUG=True
+LOG_LEVEL=INFO
+STREAMLIT_SERVER_PORT=8501
+STREAMLIT_SERVER_ADDRESS=localhost
+
+# File Upload Settings
+MAX_FILE_SIZE_MB=10
+SUPPORTED_FILE_TYPES=pdf,docx,txt
+
+# AI Model Settings
+TEMPERATURE=0.2
+MAX_TOKENS=2048
+EMBEDDING_MODEL=models/text-embedding-004
+
+# Storage Configuration
+CHROMA_PERSIST_DIRECTORY=./data/chroma_db
+UPLOAD_DIR=./uploads
+DATA_DIR=./data
+LOG_FILE=./data/app.log
+
+# Security Settings
+SECRET_KEY=your-secret-key-here
+SESSION_TIMEOUT_MINUTES=60
+```
+
+### 5. **Run the application**:
+
+```bash
+# If using UV (recommended)
+uv run streamlit run main.py
+
+# Or with activated virtual environment
+streamlit run main.py
+```
+
+### 6. **Open your browser** to `http://localhost:8501`
+
+### 🎯 Try the Demo
+
+Once running, you can immediately test the application with the included sample documents:
+
+- Navigate to **📄 Upload** page
+- Try the sample documents: Employment contracts, NDAs, Lease agreements, Service agreements
+- Experience the full analysis workflow without needing your own documents
+
+## 🐳 Docker Deployment
+
+### Local Docker Deployment
+
+```bash
+# Build the Docker image
+docker build -t lega-ai .
+
+# Run the container
+docker run -p 7860:7860 -e GOOGLE_API_KEY=your_api_key_here lega-ai
+```
+
+### Hugging Face Spaces Deployment
+
+Deploy Lega.AI to Hugging Face Spaces with one click!
+
+[![Deploy to Hugging Face Spaces](https://huggingface.co/datasets/huggingface/badges/raw/main/deploy-to-spaces-md.svg)](https://huggingface.co/spaces)
+
+**Quick Setup:**
+
+1. Create a new [Hugging Face Space](https://huggingface.co/spaces) with SDK: Docker
+2. Upload this repository to your Space
+3. Set `GOOGLE_API_KEY` in Space Settings → Variables
+4. Your app will be live at `https://huggingface.co/spaces/[username]/[space-name]`
+
+📋 **Detailed Instructions**: See [HUGGINGFACE_DEPLOYMENT.md](./HUGGINGFACE_DEPLOYMENT.md) for complete setup guide.
+
+## 📁 Project Structure
+
+```
+Lega.AI/
+├── main.py                 # Main Streamlit application entry point
+├── pyproject.toml          # UV/pip package configuration and dependencies
+├── requirements.txt        # Docker-compatible requirements file
+├── uv.lock                 # UV lockfile for reproducible builds
+├── setup.py                # Legacy Python package setup
+├── Dockerfile              # Docker container configuration
+├── .dockerignore          # Docker build optimization
+├── start.sh               # Hugging Face Spaces startup script
+├── .env.example           # Environment variables template
+├── .env.hf                # Hugging Face Spaces configuration
+├── README.md              # Project documentation
+├── HUGGINGFACE_DEPLOYMENT.md # HF Spaces deployment guide
+├── src/                   # Main application source code
+│   ├── __init__.py
+│   ├── models/
+│   │   ├── __init__.py
+│   │   └── document.py    # Document data models and schemas
+│   ├── services/
+│   │   ├── __init__.py
+│   │   ├── document_processor.py  # PDF/DOCX text extraction
+│   │   ├── ai_analyzer.py         # AI analysis and risk assessment
+│   │   └── vector_store.py        # Chroma vector database management
+│   ├── pages/
+│   │   ├── __init__.py
+│   │   ├── upload.py      # Document upload interface
+│   │   ├── analysis.py    # Document analysis dashboard
+│   │   ├── qa_assistant.py # Interactive Q&A chat interface
+│   │   ├── library.py     # Document library management
+│   │   └── settings.py    # Application settings and configuration
+│   └── utils/
+│       ├── __init__.py
+│       ├── config.py      # Environment configuration management
+│       ├── logger.py      # Logging utilities and setup
+│       └── helpers.py     # Common helper functions
+├── sample/                # Sample legal documents for testing
+│   ├── Employment_Offer_Letter.pdf
+│   ├── Master_Services_Agreement.pdf
+│   ├── Mutual_NDA.pdf
+│   └── Residential_Lease_Agreement.pdf
+├── data/                  # Local data storage and persistence
+│   ├── app.log           # Application logs
+│   └── chroma_db/        # Vector database storage
+└── uploads/              # Temporary file uploads directory
+```
+
+## 🎯 Usage Guide
+
+### 1. Document Upload & Processing
+
+- Navigate to **📄 Upload** page
+- Upload PDF, DOCX, or TXT files (max 10MB per file)
+- Try the included sample documents for immediate testing
+- Automatic document type detection and text extraction
+
+### 2. Comprehensive Analysis Dashboard
+
+Visit **📊 Analysis** to explore:
+
+- **Risk Score Gauge**: Interactive 0-100 risk assessment with color coding
+- **Side-by-Side Comparison**: Original text vs. simplified plain language
+- **Risk Factor Breakdown**: Detailed explanations of identified risks with severity levels
+- **Interactive Clause Highlighting**: Hover over highlighted text for tooltips with suggestions
+- **Financial & Date Extraction**: Automatic identification of monetary amounts and key dates
+- **Risk Visualization Charts**: Visual distribution of risk categories and severity
+
+### 3. Interactive Q&A Assistant
+
+- Use **💬 Q&A** for document-specific questions and analysis
+- Get context-aware answers powered by vector similarity search
+- Access suggested questions based on document type and content
+- Chat history preservation for reference and record-keeping
+
+### 4. Document Library Management
+
+- **📚 Library** provides persistent storage of all analyzed documents
+- Advanced filtering by document type, risk level, upload date
+- Full-text search across document content and analysis results
+- Quick re-analysis and direct access to Q&A for stored documents
+- Document metadata and analysis summary views
+
+### 5. Settings & Configuration
+
+- **⚙️ Settings** for API key management and validation
+- Application configuration and performance monitoring
+- Usage statistics and system health information
+
+## 🔧 Configuration Options
+
+The application uses environment variables for configuration. All settings can be customized in the `.env` file based on the `.env.example` template.
+
+### 🔑 Required Settings
+
+| Variable         | Description                      | Example                       |
+| ---------------- | -------------------------------- | ----------------------------- |
+| `GOOGLE_API_KEY` | Google AI API key for Gemini Pro | `xyz` (from AI Studio) |
+
+### ⚙️ Application Settings
+
+| Variable                   | Default        | Description                        |
+| -------------------------- | -------------- | ---------------------------------- |
+| `DEBUG`                    | `True`         | Enable debug mode and verbose logs |
+| `LOG_LEVEL`                | `INFO`         | Logging level (DEBUG/INFO/WARNING) |
+| `STREAMLIT_SERVER_PORT`    | `8501`         | Port for Streamlit server          |
+| `STREAMLIT_SERVER_ADDRESS` | `localhost`    | Server address binding             |
+| `MAX_FILE_SIZE_MB`         | `10`           | Maximum upload file size           |
+| `SUPPORTED_FILE_TYPES`     | `pdf,docx,txt` | Allowed file extensions            |
+
+### 🤖 AI Model Settings
+
+| Variable          | Default                | Description                      |
+| ----------------- | ---------------------- | -------------------------------- |
+| `TEMPERATURE`     | `0.2`                  | AI response creativity (0.0-1.0) |
+| `MAX_TOKENS`      | `2048`                 | Maximum response length          |
+| `EMBEDDING_MODEL` | `models/embedding-001` | Google AI embedding model        |
+
+### 💾 Storage Configuration
+
+| Variable                   | Default            | Description                  |
+| -------------------------- | ------------------ | ---------------------------- |
+| `CHROMA_PERSIST_DIRECTORY` | `./data/chroma_db` | Vector database storage path |
+| `UPLOAD_DIR`               | `./uploads`        | Temporary file uploads       |
+| `DATA_DIR`                 | `./data`           | Application data directory   |
+| `LOG_FILE`                 | `./data/app.log`   | Application log file path    |
+
+### 🔒 Security Settings
+
+| Variable                  | Default | Description              |
+| ------------------------- | ------- | ------------------------ |
+| `SECRET_KEY`              | None    | Application secret key   |
+| `SESSION_TIMEOUT_MINUTES` | `60`    | Session timeout duration |
+
+### Example .env configuration:
+
+```bash
+# Required
+GOOGLE_API_KEY=your-google-ai-api-key
+
+# Optional (with defaults shown)
+DEBUG=True
+LOG_LEVEL=INFO
+MAX_FILE_SIZE_MB=10
+SUPPORTED_FILE_TYPES=pdf,docx,txt
+CHROMA_PERSIST_DIRECTORY=./data/chroma_db
+TEMPERATURE=0.2
+```
+
+## � Sample Documents
+
+The project includes professionally-crafted sample legal documents for testing and demonstration:
+
+| Document Type                | Filename                          | Purpose                                  |
+| ---------------------------- | --------------------------------- | ---------------------------------------- |
+| **Employment Contract**      | `Employment_Offer_Letter.pdf`     | Test employment-related clause analysis  |
+| **Service Agreement**        | `Master_Services_Agreement.pdf`   | Demonstrate commercial contract analysis |
+| **Non-Disclosure Agreement** | `Mutual_NDA.pdf`                  | Show confidentiality clause assessment   |
+| **Lease Agreement**          | `Residential_Lease_Agreement.pdf` | Test rental/property contract analysis   |
+
+These documents are located in the `sample/` directory and can be uploaded directly through the application to:
+
+- Experience the complete analysis workflow
+- Test different document types and complexity levels
+- Understand risk assessment capabilities
+- Explore Q&A functionality with real legal content
+
+## �🚨 Document Types Supported
+
+Currently optimized for:
+
+- **🏠 Rental/Lease Agreements**
+- **💰 Loan Contracts**
+- **💼 Employment Contracts**
+- **🤝 Service Agreements**
+- **🔒 Non-Disclosure Agreements (NDAs)**
+- **📄 General Legal Documents**
+
+## ⚡ Key Features Deep Dive
+
+### 🔍 Advanced Risk Assessment Engine
+
+- **Multi-dimensional Analysis**: Evaluates financial, legal commitment, and rights-related risks
+- **Intelligent Severity Classification**: Categorizes risks as Low, Medium, High, or Critical
+- **Contextual Risk Scoring**: Dynamic 0-100 scale based on document type and complexity
+- **Actionable Recommendations**: Specific suggestions for improving problematic clauses
+
+### 📝 AI-Powered Plain Language Translation
+
+- **Context-Aware Simplification**: Maintains legal accuracy while improving readability
+- **Jargon Definition System**: Interactive tooltips for complex legal terms
+- **Document Type Optimization**: Tailored simplification based on contract category
+- **Preservation of Legal Intent**: Ensures meaning is not lost in translation
+
+### 🎯 Interactive Clause Analysis
+
+- **Smart Highlighting System**: Visual identification of risky and important clauses
+- **Hover Tooltips**: Immediate access to explanations and suggestions
+- **Clause Categorization**: Organized by risk type and legal significance
+- **Improvement Suggestions**: Specific recommendations for clause modifications
+
+### 🔍 Vector-Powered Document Intelligence
+
+- **Semantic Search**: Find similar clauses across your document library
+- **Context-Aware Q&A**: Answers grounded in actual document content
+- **Document Similarity**: Compare clauses against known patterns and standards
+- **Persistent Knowledge Base**: Chroma vector database for fast, accurate retrieval
+
+### 📊 Advanced Visualization & Analytics
+
+- **Interactive Risk Gauges**: Real-time visual risk assessment
+- **Risk Distribution Charts**: Breakdown of risk categories and severity
+- **Financial Terms Extraction**: Automatic identification of monetary obligations
+- **Timeline Analysis**: Key dates and deadline extraction with visualization
+
+### 💾 Enterprise-Grade Data Management
+
+- **Local Data Persistence**: Secure storage of documents and analysis results
+- **Document Library**: Organized management with search and filtering
+- **Analysis History**: Complete audit trail of document processing
+- **Metadata Extraction**: Automatic tagging and categorization
+
+## 🔒 Privacy & Security
+
+### 🛡️ Data Protection
+
+- **Local Processing**: Documents analyzed locally with secure API calls to Google AI
+- **No Data Sharing**: Zero third-party data sharing or storage outside your environment
+- **Secure Storage**: Vector embeddings and analysis results stored locally in Chroma database
+- **Environment Security**: API keys managed through secure environment variables
+
+### 🔐 Security Best Practices
+
+- **API Key Protection**: Secure credential management with environment-based configuration
+- **Local Vector Storage**: Document embeddings stored exclusively on your local system
+- **Session Management**: Configurable session timeouts and secure state management
+- **Input Validation**: Comprehensive file type and size validation for uploads
+
+### 📋 Data Handling
+
+- **Temporary Upload Storage**: Uploaded files processed and optionally removed from temp storage
+- **Persistent Analysis**: Analysis results retained locally for document library functionality
+- **User Control**: Complete control over data retention and deletion
+- **Audit Trail**: Transparent logging of all document processing activities
+
+## 🤝 Contributing
+
+1. Fork the repository
+2. Create a feature branch
+3. Make your changes
+4. Test thoroughly
+5. Submit a pull request
+
+## 📄 License
+
+MIT License - see LICENSE file for details.
+
+## 🆘 Support
+
+### 📚 Documentation & Resources
+
+- **In-Code Documentation**: Comprehensive docstrings and code comments throughout the project
+- **Configuration Guide**: Detailed environment setup and configuration options above
+- **Sample Documents**: Use included sample contracts to understand features and capabilities
+
+### 🐛 Issues & Bug Reports
+
+- **GitHub Issues**: Report bugs, request features, or ask questions via [GitHub Issues](https://github.com/your-repo/Lega.AI/issues)
+- **Bug Reports**: Include system info, error logs, and steps to reproduce
+- **Feature Requests**: Describe use cases and expected functionality
+
+### 🛠️ Development & API References
+
+- **Google AI Documentation**: [Google AI Developer Guide](https://ai.google.dev/) for Gemini API details
+- **LangChain Documentation**: [LangChain Docs](https://python.langchain.com/) for framework reference
+- **Streamlit Documentation**: [Streamlit Docs](https://docs.streamlit.io/) for UI framework guidance
+- **Chroma Documentation**: [Chroma Docs](https://docs.trychroma.com/) for vector database operations
+
+### 💡 Getting Help
+
+1. **Check Documentation**: Review this README and in-code comments first
+2. **Try Sample Documents**: Use provided samples to test functionality
+3. **Check Logs**: Review `data/app.log` for detailed error information
+4. **Environment Issues**: Verify `.env` configuration and API key validity
+5. **Community Support**: Open GitHub discussions for general questions
+
+---
+
+**Made with ❤️ using Streamlit, LangChain, and Google AI**

main.py

@@ -0,0 +1,789 @@
+import streamlit as st
+from streamlit_option_menu import option_menu
+import os
+import sys
+
+# Add src directory to Python path
+sys.path.append(os.path.join(os.path.dirname(__file__), "src"))
+
+from src.utils.config import config
+from src.utils.logger import setup_logging
+
+# Page config
+st.set_page_config(
+    page_title="Lega.AI", page_icon="⚖️", layout="wide", initial_sidebar_state="expanded"
+)
+
+# Custom CSS for responsive dark/light theme
+st.markdown(
+    """
+<style>
+    /* Main header with gradient text */
+    .main-header {
+        font-size: 3rem;
+        font-weight: bold;
+        text-align: center;
+        margin-bottom: 2rem;
+        background: linear-gradient(90deg, #1f4e79, #2e86ab);
+        -webkit-background-clip: text;
+        -webkit-text-fill-color: transparent;
+        background-clip: text;
+    }
+    
+    /* Responsive feature cards that adapt to theme */
+    .feature-card {
+        background: var(--background-color);
+        padding: 1.5rem;
+        border-radius: 12px;
+        box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1);
+        margin: 1rem 0;
+        border-left: 4px solid #2e86ab;
+        border: 1px solid var(--border-color);
+        color: var(--text-color);
+        transition: all 0.3s ease;
+    }
+    
+    .feature-card:hover {
+        transform: translateY(-2px);
+        box-shadow: 0 6px 12px rgba(0, 0, 0, 0.15);
+    }
+    
+    /* Risk color indicators */
+    .risk-critical { color: #ff4444; font-weight: bold; }
+    .risk-high { color: #ff6666; font-weight: bold; }
+    .risk-medium { color: #ffaa00; font-weight: bold; }
+    .risk-low { color: #ffcc00; font-weight: bold; }
+    .risk-safe { color: #44aa44; font-weight: bold; }
+    
+    /* Responsive metric cards */
+    .metric-card {
+        background: var(--secondary-background-color);
+        padding: 1.5rem;
+        border-radius: 10px;
+        text-align: center;
+        margin: 0.5rem 0;
+        border: 1px solid var(--border-color);
+        color: var(--text-color);
+        transition: all 0.3s ease;
+    }
+    
+    .metric-card:hover {
+        background: var(--hover-background-color);
+    }
+    
+    /* Enhanced button styling */
+    .stButton > button {
+        width: 100%;
+        background: linear-gradient(135deg, #2e86ab, #1f4e79);
+        color: white !important;
+        border: none;
+        border-radius: 10px;
+        padding: 0.75rem 1rem;
+        font-weight: 600;
+        transition: all 0.3s ease;
+        box-shadow: 0 2px 4px rgba(0, 0, 0, 0.2);
+    }
+    
+    .stButton > button:hover {
+        background: linear-gradient(135deg, #1f4e79, #2e86ab);
+        transform: translateY(-1px);
+        box-shadow: 0 4px 8px rgba(0, 0, 0, 0.3);
+    }
+    
+    /* Enhanced sidebar styling for dark/light theme */
+    .css-1d391kg {
+        background: var(--background-color) !important;
+    }
+    
+    /* Streamlit sidebar container */
+    section[data-testid="stSidebar"] {
+        background: var(--background-color) !important;
+        border-right: 1px solid var(--border-color) !important;
+    }
+    
+    /* Sidebar content */
+    section[data-testid="stSidebar"] > div {
+        background: var(--background-color) !important;
+        color: var(--text-color) !important;
+    }
+    
+    /* Sidebar header */
+    section[data-testid="stSidebar"] .block-container {
+        background: var(--background-color) !important;
+        color: var(--text-color) !important;
+    }
+    
+    /* Option menu in sidebar */
+    section[data-testid="stSidebar"] .nav-link {
+        background: var(--secondary-background-color) !important;
+        color: var(--text-color) !important;
+        border: 1px solid var(--border-color) !important;
+    }
+    
+    /* Active option in sidebar */
+    section[data-testid="stSidebar"] .nav-link.active {
+        background: linear-gradient(135deg, #2e86ab, #1f4e79) !important;
+        color: white !important;
+    }
+    
+    /* Streamlit Option Menu specific styling - Force override */
+    .nav-link {
+        background: var(--secondary-background-color) !important;
+        color: var(--text-color) !important;
+        border: 1px solid var(--border-color) !important;
+        border-radius: 8px !important;
+        margin: 2px 0 !important;
+    }
+    
+    .nav-link:hover {
+        background: var(--hover-background-color) !important;
+        color: var(--text-color) !important;
+    }
+    
+    .nav-link.active {
+        background: linear-gradient(135deg, #2e86ab, #1f4e79) !important;
+        color: white !important;
+        border: 1px solid #2e86ab !important;
+    }
+    
+    /* Option menu container */
+    .nav {
+        background: transparent !important;
+    }
+    
+    /* Fix option menu wrapper */
+    div[data-testid="stVerticalBlock"] > div > div {
+        background: transparent !important;
+    }
+    
+    /* More specific selectors for option menu */
+    section[data-testid="stSidebar"] .nav-link {
+        background-color: var(--secondary-background-color) !important;
+        color: var(--text-color) !important;
+        border: 1px solid var(--border-color) !important;
+    }
+    
+    section[data-testid="stSidebar"] .nav-link:hover {
+        background-color: var(--hover-background-color) !important;
+    }
+    
+    section[data-testid="stSidebar"] .nav-link.active {
+        background-color: #2e86ab !important;
+        color: white !important;
+    }
+    
+    /* Force override any white backgrounds in sidebar */
+    section[data-testid="stSidebar"] * {
+        background-color: inherit !important;
+    }
+    
+    section[data-testid="stSidebar"] .nav-link {
+        background-color: var(--secondary-background-color) !important;
+    }
+    
+    /* File uploader styling */
+    .uploadedFile {
+        background: var(--secondary-background-color) !important;
+        border: 2px dashed var(--border-color) !important;
+        border-radius: 10px !important;
+        color: var(--text-color) !important;
+    }
+    
+    /* Tab styling */
+    .stTabs [data-baseweb="tab-list"] {
+        gap: 8px;
+    }
+    
+    .stTabs [data-baseweb="tab"] {
+        background: var(--secondary-background-color);
+        border-radius: 8px;
+        color: var(--text-color);
+        border: 1px solid var(--border-color);
+    }
+    
+    .stTabs [aria-selected="true"] {
+        background: linear-gradient(135deg, #2e86ab, #1f4e79) !important;
+        color: white !important;
+    }
+    
+    /* Tooltip styling for risk factors */
+    .tooltip {
+        position: relative;
+        display: inline;
+        cursor: help;
+        border-radius: 4px;
+        padding: 2px 4px;
+        margin: 0 1px;
+    }
+    
+    .tooltip .tooltiptext {
+        visibility: hidden;
+        width: 300px;
+        background-color: var(--tooltip-background);
+        color: var(--tooltip-text);
+        text-align: left;
+        border-radius: 8px;
+        padding: 12px;
+        position: absolute;
+        z-index: 1000;
+        bottom: 125%;
+        left: 50%;
+        margin-left: -150px;
+        opacity: 0;
+        transition: opacity 0.3s, visibility 0.3s;
+        font-size: 13px;
+        line-height: 1.4;
+        box-shadow: 0 4px 12px rgba(0, 0, 0, 0.3);
+        border: 1px solid var(--border-color);
+    }
+    
+    .tooltip:hover .tooltiptext {
+        visibility: visible;
+        opacity: 1;
+    }
+    
+    /* Risk highlighting */
+    .risk-critical { 
+        background-color: rgba(255, 68, 68, 0.2); 
+        border-left: 4px solid #ff4444; 
+        padding: 4px 8px;
+        border-radius: 4px;
+    }
+    .risk-high { 
+        background-color: rgba(255, 136, 0, 0.2); 
+        border-left: 4px solid #ff8800; 
+        padding: 4px 8px;
+        border-radius: 4px;
+    }
+    .risk-medium { 
+        background-color: rgba(255, 204, 0, 0.2); 
+        border-left: 4px solid #ffcc00; 
+        padding: 4px 8px;
+        border-radius: 4px;
+    }
+    .risk-low { 
+        background-color: rgba(68, 170, 68, 0.2); 
+        border-left: 4px solid #44aa44; 
+        padding: 4px 8px;
+        border-radius: 4px;
+    }
+    
+    /* Jargon term highlighting */
+    .jargon-term { 
+        background-color: rgba(46, 134, 171, 0.2); 
+        text-decoration: underline dotted #2e86ab;
+        padding: 2px 4px;
+        border-radius: 3px;
+    }
+    
+    /* Dark theme variables */
+    [data-theme="dark"] {
+        --background-color: #0e1117;
+        --secondary-background-color: #262730;
+        --text-color: #fafafa;
+        --border-color: #464a5a;
+        --hover-background-color: #3d4354;
+        --tooltip-background: #262730;
+        --tooltip-text: #fafafa;
+    }
+    
+    /* Light theme variables */
+    [data-theme="light"], :root {
+        --background-color: #ffffff;
+        --secondary-background-color: #f8f9fa;
+        --text-color: #262626;
+        --border-color: #e0e0e0;
+        --hover-background-color: #f0f0f0;
+        --tooltip-background: #333333;
+        --tooltip-text: #ffffff;
+    }
+    
+    /* Auto-detect system theme */
+    @media (prefers-color-scheme: dark) {
+        :root {
+            --background-color: #0e1117;
+            --secondary-background-color: #262730;
+            --text-color: #fafafa;
+            --border-color: #464a5a;
+            --hover-background-color: #3d4354;
+            --tooltip-background: #262730;
+            --tooltip-text: #fafafa;
+        }
+    }
+    
+    /* Hide Streamlit default elements */
+    #MainMenu {visibility: hidden;}
+    footer {visibility: hidden;}
+    .stDeployButton {visibility: hidden;}
+    
+    /* Fix all Streamlit components for dark theme */
+    .stApp {
+        background: var(--background-color) !important;
+        color: var(--text-color) !important;
+    }
+    
+    /* Main content area */
+    .main .block-container {
+        background: var(--background-color) !important;
+        color: var(--text-color) !important;
+        padding-top: 2rem;
+        padding-bottom: 2rem;
+    }
+    
+    /* Text input fields */
+    .stTextInput > div > div > input {
+        background: var(--secondary-background-color) !important;
+        color: var(--text-color) !important;
+        border: 1px solid var(--border-color) !important;
+    }
+    
+    /* Text areas */
+    .stTextArea > div > div > textarea {
+        background: var(--secondary-background-color) !important;
+        color: var(--text-color) !important;
+        border: 1px solid var(--border-color) !important;
+    }
+    
+    /* Select boxes */
+    .stSelectbox > div > div > select {
+        background: var(--secondary-background-color) !important;
+        color: var(--text-color) !important;
+        border: 1px solid var(--border-color) !important;
+    }
+    
+    /* Info, warning, error boxes */
+    .stAlert {
+        background: var(--secondary-background-color) !important;
+        color: var(--text-color) !important;
+        border: 1px solid var(--border-color) !important;
+    }
+    
+    /* Expander */
+    .streamlit-expanderHeader {
+        background: var(--secondary-background-color) !important;
+        color: var(--text-color) !important;
+        border: 1px solid var(--border-color) !important;
+    }
+    
+    /* Columns */
+    .element-container {
+        background: transparent !important;
+    }
+    
+    /* Status indicators */
+    .status-success {
+        background: rgba(68, 170, 68, 0.1);
+        border: 1px solid #44aa44;
+        border-radius: 6px;
+        padding: 8px 12px;
+        color: #44aa44;
+    }
+    
+    .status-warning {
+        background: rgba(255, 136, 0, 0.1);
+        border: 1px solid #ff8800;
+        border-radius: 6px;
+        padding: 8px 12px;
+        color: #ff8800;
+    }
+    
+    .status-error {
+        background: rgba(255, 68, 68, 0.1);
+        border: 1px solid #ff4444;
+        border-radius: 6px;
+        padding: 8px 12px;
+        color: #ff4444;
+    }
+</style>
+""",
+    unsafe_allow_html=True,
+)
+
+
+def main():
+    # Initialize logging
+    setup_logging()
+
+    # Initialize session state
+    if "current_document" not in st.session_state:
+        st.session_state.current_document = None
+    if "documents_library" not in st.session_state:
+        st.session_state.documents_library = []
+
+    # Sidebar navigation
+    with st.sidebar:
+        st.markdown("### ⚖️ Lega.AI")
+        st.markdown("*Making legal documents accessible*")
+
+        selected = option_menu(
+            menu_title=None,
+            options=[
+                "🏠 Home",
+                "📄 Upload",
+                "📊 Analysis",
+                "💬 Q&A",
+                "📚 Library",
+                "⚙️ Settings",
+            ],
+            icons=["house", "upload", "graph-up", "chat-dots", "folder", "gear"],
+            menu_icon="list",
+            default_index=0,
+            styles={
+                "container": {
+                    "padding": "0!important",
+                    "background-color": "transparent",
+                },
+                "icon": {"color": "#2e86ab", "font-size": "18px"},
+                "nav-link": {
+                    "font-size": "16px",
+                    "text-align": "left",
+                    "margin": "2px 0px",
+                    "padding": "8px 12px",
+                    "border-radius": "8px",
+                    "background-color": "transparent",
+                    "color": "inherit",
+                    "border": "1px solid transparent",
+                    "--hover-color": "transparent",
+                },
+                "nav-link-selected": {
+                    "background-color": "#2e86ab",
+                    "color": "white",
+                    "border": "1px solid #2e86ab",
+                },
+            },
+        )
+
+    # Handle page redirections from session state
+    if "page" in st.session_state and st.session_state.page:
+        # Map the session state page to the selected value
+        page_mapping = {
+            "📄 Upload": "📄 Upload",
+            "📊 Analysis": "📊 Analysis",
+            "💬 Q&A": "💬 Q&A",
+            "📚 Library": "📚 Library",
+            "⚙️ Settings": "⚙️ Settings",
+        }
+
+        if st.session_state.page in page_mapping:
+            selected = st.session_state.page
+            # Clear the page state to prevent continuous redirections
+            del st.session_state.page
+
+    # Main content area
+    if selected == "🏠 Home":
+        show_home_page()
+    elif selected == "📄 Upload":
+        show_upload_page()
+    elif selected == "📊 Analysis":
+        show_analysis_page()
+    elif selected == "💬 Q&A":
+        show_qa_page()
+    elif selected == "📚 Library":
+        show_library_page()
+    elif selected == "⚙️ Settings":
+        show_settings_page()
+
+
+def show_home_page():
+    """Display the home page with overview and features."""
+    st.markdown('<h1 class="main-header">⚖️ Lega.AI</h1>', unsafe_allow_html=True)
+    st.markdown(
+        '<p style="text-align: center; font-size: 1.2rem; color: #666;">AI-powered legal document analysis and simplification</p>',
+        unsafe_allow_html=True,
+    )
+
+    # Key benefits
+    col1, col2, col3 = st.columns(3)
+
+    with col1:
+        st.markdown(
+            """
+        <div class="feature-card">
+            <h3>🚀 Instant Analysis</h3>
+            <p>Upload any legal document and get comprehensive analysis in under 60 seconds using Google's Gemini AI.</p>
+        </div>
+        """,
+            unsafe_allow_html=True,
+        )
+
+    with col2:
+        st.markdown(
+            """
+        <div class="feature-card">
+            <h3>🎯 Risk Assessment</h3>
+            <p>Color-coded risk scoring helps you identify problematic clauses at a glance with detailed explanations.</p>
+        </div>
+        """,
+            unsafe_allow_html=True,
+        )
+
+    with col3:
+        st.markdown(
+            """
+        <div class="feature-card">
+            <h3>💬 Plain Language</h3>
+            <p>Convert complex legal jargon into clear, understandable language that anyone can comprehend.</p>
+        </div>
+        """,
+            unsafe_allow_html=True,
+        )
+
+    # Quick stats
+    st.markdown("---")
+    st.subheader("📊 Platform Statistics")
+
+    col1, col2, col3, col4 = st.columns(4)
+
+    with col1:
+        st.markdown(
+            """
+        <div class="metric-card">
+            <h2>1,247</h2>
+            <p>Documents Analyzed</p>
+        </div>
+        """,
+            unsafe_allow_html=True,
+        )
+
+    with col2:
+        st.markdown(
+            """
+        <div class="metric-card">
+            <h2>95%</h2>
+            <p>Accuracy Rate</p>
+        </div>
+        """,
+            unsafe_allow_html=True,
+        )
+
+    with col3:
+        st.markdown(
+            """
+        <div class="metric-card">
+            <h2>₹2,000</h2>
+            <p>Avg. Saved per User</p>
+        </div>
+        """,
+            unsafe_allow_html=True,
+        )
+
+    with col4:
+        st.markdown(
+            """
+        <div class="metric-card">
+            <h2>45 sec</h2>
+            <p>Avg. Processing Time</p>
+        </div>
+        """,
+            unsafe_allow_html=True,
+        )
+
+    # Getting started
+    st.markdown("---")
+    st.subheader("🎯 Getting Started")
+
+    col1, col2 = st.columns([2, 1])
+
+    with col1:
+        st.markdown(
+            """
+        **How to use Lega.AI:**
+        
+        1. **Upload** your legal document (PDF, DOCX, or TXT)
+        2. **Wait** for AI analysis (typically 30-60 seconds)
+        3. **Review** risk assessment and simplified explanations  
+        4. **Ask questions** about specific clauses or terms
+        5. **Export** summary for your records
+        """
+        )
+
+    with col2:
+        st.markdown("### 📄 Try Real Sample Documents")
+        st.markdown("Get started with actual legal documents:")
+        
+        # Get available sample documents
+        sample_dir = "./sample"
+        sample_files = []
+        if os.path.exists(sample_dir):
+            sample_files = [f for f in os.listdir(sample_dir) if f.endswith(('.pdf', '.docx', '.txt'))]
+        
+        if sample_files:
+            for i, filename in enumerate(sample_files[:4]):  # Show first 4
+                display_name = filename.replace('_', ' ').replace('.pdf', '').replace('.docx', '').replace('.txt', '')
+                display_name = display_name.title()
+                
+                if st.button(f"📄 {display_name}", key=f"home_sample_{i}"):
+                    st.session_state.load_sample = filename
+                    st.session_state.page = "📄 Upload"
+                    st.rerun()
+        else:
+            st.info("Sample documents loading...")
+
+    # CTA button
+    st.markdown("---")
+    if st.button("📄 Analyze Your First Document", type="primary"):
+        st.session_state.page = "📄 Upload"
+        st.rerun()
+
+
+def load_sample_document(doc_type: str):
+    """Load a sample document for demonstration."""
+    sample_docs = {
+        "rental": {
+            "filename": "sample_rental_agreement.pdf",
+            "type": "rental",
+            "text": """
+            RESIDENTIAL LEASE AGREEMENT
+            
+            This Lease Agreement is entered into between John Smith (Landlord) and Jane Doe (Tenant) 
+            for the property located at 123 Main Street, Mumbai, Maharashtra.
+            
+            RENT: Tenant agrees to pay Rs. 25,000 per month, due on the 1st of each month. 
+            Late payments will incur a penalty of Rs. 1,000 per day.
+            
+            SECURITY DEPOSIT: Tenant shall pay a security deposit of Rs. 75,000, which is 
+            non-refundable except for damage assessment.
+            
+            TERMINATION: Either party may terminate this lease with 30 days written notice. 
+            Early termination by Tenant results in forfeiture of security deposit.
+            
+            MAINTENANCE: Tenant is responsible for all repairs and maintenance, including 
+            structural repairs, regardless of cause.
+            
+            The property is leased "as-is" with no warranties. Landlord is not liable for 
+            any damages or injuries occurring on the premises.
+            """,
+        },
+        "loan": {
+            "filename": "sample_loan_agreement.pdf",
+            "type": "loan",
+            "text": """
+            PERSONAL LOAN AGREEMENT
+            
+            Borrower: Rajesh Kumar
+            Lender: QuickCash Financial Services Pvt Ltd
+            Principal Amount: Rs. 2,00,000
+            
+            INTEREST RATE: 24% per annum (APR 28.5% including processing fees)
+            
+            REPAYMENT: 24 monthly installments of Rs. 12,500 each
+            Total repayment amount: Rs. 3,00,000
+            
+            LATE PAYMENT PENALTY: Rs. 500 per day for any late payment
+            
+            DEFAULT: If payment is late by more than 7 days, the entire remaining 
+            balance becomes immediately due and payable.
+            
+            COLLATERAL: Borrower pledges gold ornaments worth Rs. 2,50,000 as security.
+            Lender may seize collateral immediately upon default.
+            
+            ARBITRATION: All disputes shall be resolved through binding arbitration. 
+            Borrower waives right to jury trial.
+            
+            Processing fee: Rs. 10,000 (non-refundable)
+            Documentation charges: Rs. 5,000
+            """,
+        },
+        "employment": {
+            "filename": "sample_employment_contract.pdf",
+            "type": "employment",
+            "text": """
+            EMPLOYMENT CONTRACT
+            
+            Employee: Priya Sharma
+            Company: TechCorp India Private Limited
+            Position: Software Developer
+            Start Date: January 1, 2024
+            
+            SALARY: Rs. 8,00,000 per annum, payable monthly
+            
+            WORKING HOURS: 45 hours per week, including mandatory weekend work when required
+            
+            NON-COMPETE: Employee shall not work for any competing company for 2 years 
+            after termination, within India or globally.
+            
+            CONFIDENTIALITY: Employee agrees to maintain strict confidentiality of all 
+            company information indefinitely, even after termination.
+            
+            TERMINATION: Company may terminate employment at any time without cause or notice. 
+            Employee must provide 90 days notice to resign.
+            
+            NO MOONLIGHTING: Employee shall not engage in any other work or business 
+            activities during employment.
+            
+            INTELLECTUAL PROPERTY: All work created by employee belongs entirely to company, 
+            including personal projects done outside work hours.
+            """,
+        },
+    }
+
+    if doc_type in sample_docs:
+        sample = sample_docs[doc_type]
+        from src.utils.helpers import generate_document_id
+
+        # Store in session state
+        st.session_state.current_document = {
+            "id": generate_document_id(),
+            "filename": sample["filename"],
+            "document_type": sample["type"],
+            "original_text": sample["text"],
+            "is_sample": True,
+        }
+
+        st.success(f"📄 Loaded sample {doc_type} document!")
+        st.session_state.page = "📊 Analysis"
+        st.rerun()
+
+
+def show_upload_page():
+    """Import and show the upload page."""
+    try:
+        from src.pages.upload import show_upload_interface
+
+        show_upload_interface()
+    except ImportError as e:
+        st.error(f"Upload page not found: {e}")
+
+
+def show_analysis_page():
+    """Import and show the analysis page."""
+    try:
+        from src.pages.analysis import show_analysis_interface
+
+        show_analysis_interface()
+    except ImportError as e:
+        st.error(f"Analysis page not found: {e}")
+
+
+def show_qa_page():
+    """Import and show the Q&A page."""
+    try:
+        from src.pages.qa_assistant import show_qa_interface
+
+        show_qa_interface()
+    except ImportError as e:
+        st.error(f"Q&A page not found: {e}")
+
+
+def show_library_page():
+    """Import and show the library page."""
+    try:
+        from src.pages.library import show_library_interface
+
+        show_library_interface()
+    except ImportError as e:
+        st.error(f"Library page not found: {e}")
+
+
+def show_settings_page():
+    """Import and show the settings page."""
+    try:
+        from src.pages.settings import show_settings_interface
+
+        show_settings_interface()
+    except ImportError as e:
+        st.error(f"Settings page not found: {e}")
+
+
+if __name__ == "__main__":
+    main()

pyproject.toml

@@ -0,0 +1,20 @@
+[project]
+name = "lega-ai"
+version = "0.1.0"
+description = "Add your description here"
+readme = "README.md"
+requires-python = ">=3.13"
+dependencies = [
+    "blinker>=1.9.0",
+    "langchain-chroma>=0.2.6",
+    "langchain-community>=0.3.29",
+    "langchain-google-genai>=2.1.12",
+    "langchain-text-splitters>=0.3.11",
+    "langchain[google-genai]>=0.3.27",
+    "plotly>=6.3.0",
+    "pypdf>=6.0.0",
+    "python-docx>=1.2.0",
+    "python-dotenv>=1.1.1",
+    "streamlit>=1.49.1",
+    "streamlit-option-menu>=0.4.0",
+]

requirements.txt

@@ -0,0 +1,17 @@
+# Core dependencies
+blinker>=1.9.0
+langchain-chroma>=0.2.6
+langchain-community>=0.3.29
+langchain-google-genai>=2.1.12
+langchain-text-splitters>=0.3.11
+langchain[google-genai]>=0.3.27
+plotly>=6.3.0
+pypdf>=6.0.0
+python-docx>=1.2.0
+python-dotenv>=1.1.1
+streamlit>=1.49.1
+streamlit-option-menu>=0.4.0
+
+# Additional dependencies for production deployment
+uvicorn>=0.24.0
+gunicorn>=21.2.0

sample/Employment_Offer_Letter.pdf

@@ -0,0 +1,74 @@
+%PDF-1.3
+%���� ReportLab Generated PDF document http://www.reportlab.com
+1 0 obj
+<<
+/F1 2 0 R /F2 3 0 R
+>>
+endobj
+2 0 obj
+<<
+/BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font
+>>
+endobj
+3 0 obj
+<<
+/BaseFont /Times-Roman /Encoding /WinAnsiEncoding /Name /F2 /Subtype /Type1 /Type /Font
+>>
+endobj
+4 0 obj
+<<
+/Contents 8 0 R /MediaBox [ 0 0 595.2756 841.8898 ] /Parent 7 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+5 0 obj
+<<
+/PageMode /UseNone /Pages 7 0 R /Type /Catalog
+>>
+endobj
+6 0 obj
+<<
+/Author (anonymous) /CreationDate (D:20250921125755+00'00') /Creator (ReportLab PDF Library - www.reportlab.com) /Keywords () /ModDate (D:20250921125755+00'00') /Producer (ReportLab PDF Library - www.reportlab.com) 
+  /Subject (unspecified) /Title (untitled) /Trapped /False
+>>
+endobj
+7 0 obj
+<<
+/Count 1 /Kids [ 4 0 R ] /Type /Pages
+>>
+endobj
+8 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 1799
+>>
+stream
+GarVO;01G^%"@ZKd,a<:)p018Rl9XI9h<;\`6!Pa:5@n=0`ndrH[WNl;#^EA=rZ4XRnOH2<mmmAT;3LThtS(C^M-I8-)&(k7l7fMA5/hC_fq>=Bp--Gqb_G,rT3P>lJcPH+$RC\Nik"<.*5n>6).:2H98bNU[uU`_t<rkh'>bZrq:LjUZj$q-F;<gr1`7JP2GlM[E2'AltOW!Ze_oj4WPQQ;4^#"6N3O'_913c=qrR$_Ntt;S=^!U7t*-@b>`)T*3[F&inp>cn7cej;?`K]RP"LOk@J9!3>\:<`ff5rc8b4!O[HL#BHEaYVu;obgkUFC+[/&\Mre+5mFgOc+XL@[-k#:%A+.?#4G3u@(TB#e7ee`_!48KM+<ua(,VEKY*k($)I\S&Lgs;`@\rUPFhpBWW,aus31ZpubUYDe"6X9Ks10cJn!Eh%+6l-g$@kk/Z*7E-bV]g2KCP?l^<L3"S\s:S5#!h[6j2e#*QBSAt/ZgM+fh!Tc5-:>6\[#D_SMR67WE(m1iHWPN3"Kp9Lp9*%_UAmuF0u7#j\2A6[;3H)NlegKdPuOD,FUs5XUpQ'knDf^]0u%NJ$[U,2E7Jf<Wc1DWXp#i'C,Wn[*r^LJ4or7%<WgtP],Kl%<`"d3E)WY>?qm>M5Afa%r+Sl;uW9/e$Z1qbDY"tS`51#QG5EP;CY\4R8/l(;l4`Y=hc!*p&3^a,bC&c(+8o1d0J`3VX2#i3/V@U)!@bGr0%g0W/<&H:oX.3L.612eC^$9;QbTE1B5866#5DZbL`an:qKn%_PRHF84SS.Bs:K$HD:@7IDe,=\MQ`TpB1G"Bb.U_W'[,22H[2bqZdQ4qAe+bO-t8PUb@pVEZ6>)gECumi4+E[0>Z2E@>7XqC[GJ\$Z?q[D(s)/=1]J69:?BUCU6&fqd/P)haTO9D)@hRaj8>m;>a^+grI)J`o*hc)TT1#&@>VRaM0a;)KBr24BR4b`*O]$'rXg/g4#!<BD<%OG%?Y$Ws(AA2jV"7;Kn"$JjD*Q6<$fs3%U]u]oo\o8k[PBmR&h$b442d2HW=/Ea.jp**&Ct!?%1de@J"qbi.ZFAs7KqMlr9-qZM:u]5cUZIo6@V>63P3N`$A+WJZs__GR5]A1G\^N.V/.Pa)UL`eYlanlAIfWRE<'\`4*.q!i`RnbAPn;u!J2K.Jgj0SQ"lJOJd"TlCDaJ[C-80^7M!(Pemh>Uao<"t1Y4RcJ?WT#fH=29>F<T#S@f0OMOJ5?ObM7RFoqCqC;l28D/S5KirEo+^?B"tVZI'A>mfYh-^!/9X'JF."%C9bXs9"2t2KA08$.-aUE55+,`dJtfR'(>O!lG%uG:(gu+@YWQ]30&udATP6Y)!jT?cf1m6[6>s`g#3e)sGS=+\b//dsj*<"83/6.#M'XC`?FSoGKrf]>+u\rfXOBQ".9^e)ML@jQIdVd&-I"#E:RA:HF6nNs37PR!S'<mfDl"mh8tsVYZEGg*'H^OakQph2rc!X+3sJj<adc%lO$ic$NgRd_r0=/>+.m70q[t3]'BYL=l#'C:bVRuTMgW&GeO4I8ARV$SR@1o/\eS?`>%&'HiC^^[jpu>3:?4$C%`^h6[l;0QX=kj4&g*EVYKr4eBoC.i[/=1h>(a:8Yd'FQJQ$-Hg34@!L/n5DEp5<pfSXj%<CJK]7*rRnJ8hC>Ceq3?&)g8$H81\4_`\HE'!B*?4D;+9?NQ5r_>jXflLKWVPcENJVjgm7Nm!-c7EW-&i@^T;o8a<HS[)[-@[a)>(H9UKm5=Afg,Z=(8.QB/oskA&ea>0>53ne[iMTcW!WT#&Jc~>endstream
+endobj
+xref
+0 9
+0000000000 65535 f 
+0000000073 00000 n 
+0000000114 00000 n 
+0000000221 00000 n 
+0000000330 00000 n 
+0000000533 00000 n 
+0000000601 00000 n 
+0000000897 00000 n 
+0000000956 00000 n 
+trailer
+<<
+/ID 
+[<cc2bcd5827912acf8c3bd2324dddcc13><cc2bcd5827912acf8c3bd2324dddcc13>]
+% ReportLab generated PDF document -- digest (http://www.reportlab.com)
+
+/Info 6 0 R
+/Root 5 0 R
+/Size 9
+>>
+startxref
+2846
+%%EOF

sample/Master_Services_Agreement.pdf

@@ -0,0 +1,74 @@
+%PDF-1.3
+%���� ReportLab Generated PDF document http://www.reportlab.com
+1 0 obj
+<<
+/F1 2 0 R /F2 3 0 R
+>>
+endobj
+2 0 obj
+<<
+/BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font
+>>
+endobj
+3 0 obj
+<<
+/BaseFont /Times-Roman /Encoding /WinAnsiEncoding /Name /F2 /Subtype /Type1 /Type /Font
+>>
+endobj
+4 0 obj
+<<
+/Contents 8 0 R /MediaBox [ 0 0 595.2756 841.8898 ] /Parent 7 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+5 0 obj
+<<
+/PageMode /UseNone /Pages 7 0 R /Type /Catalog
+>>
+endobj
+6 0 obj
+<<
+/Author (anonymous) /CreationDate (D:20250921125755+00'00') /Creator (ReportLab PDF Library - www.reportlab.com) /Keywords () /ModDate (D:20250921125755+00'00') /Producer (ReportLab PDF Library - www.reportlab.com) 
+  /Subject (unspecified) /Title (untitled) /Trapped /False
+>>
+endobj
+7 0 obj
+<<
+/Count 1 /Kids [ 4 0 R ] /Type /Pages
+>>
+endobj
+8 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 2201
+>>
+stream
+Gas1`d;ms^&q/Z@kXsCVZ<MtkG*6-tT+i,(WX-Ebg"o^(AWR*PM56h7K*h;<n*,sr+SNL6`$Wo+GG!@?VjfES_k446R?^14;Yous9]HLQf/.q@H;A6C,5h\ka6h7b+%IEV0!<4dfA@Z>j/\5BPa8h$60#5bF?-qj97pB/UgcoP=PefHO\g#/'OiEKoQNpB;>LOf^HoHY29f\!Pbge$B:Z#jJg[TuW@CE!K/^+^Gk:\li><UaX)cXR1F-`kCN:@_@_Gg1e'EoKdOA??86.aQPhR]gdA5-h3k?:js/pK4$Yu>V%aU1<-*n%PHF&^a*T$QiR44$5E,kB-r(F/,KJPXk0cj1IR4e]fD`YPeM89R<ogT]#D3\BG>C+Kc&UCPEbj(Gu8Z#Br/="4)8fY&i,p)c4Z/:t=AH'C"W#mMTh!H)!mj`MA4Ln$;,/Nbf](N2+bh0&&bSU-]Y@YmiHfGOl`JDD+IND8$q"C-l_,jBA&("8ANj.U'-,8eVXamTUNe#i#(`tu-gMNW?\LJ;K2kmpa:c1SU41j*%W7e1EYBB#%j[3e]1iu9mPOB4Q21fJ9)ArZ<Qg_^`/e9hAXXW5tfl0VF0TRVkjIM,SEg^JSFu39'n#f.%3$0UT"tX.r4SnOtk7=Pk!6$bp9DLrA]95:WG["q'WB`T^EdpjJ,nl`)-k_a9/r42G7.i+[Y<AeMUG:t2Dq,A(4'b"C7hCuP7GT<>(MXo"<DISJlmlZ5NS\]7)S!5R"rN[[ka_j'ene@L%$jXO?SqG0m,bk']oj&NrTQsS<h-VWbDbJ[i_HRjrD_=#L2F!0a4iW':E)5<V]G?S4m0q"cFrV8n)^Q7$X,0FqH+1EX;oT[AZ=VgIH"sY4m]VSU@k[iOhUXH<:<lF5QTW)e\N%prc[V,KJd#qH8t$GOs8ha!;_o'^5lS%D7E(R;Iq>^SJAaTnIB8OCm"1+Ak:JmADn#:f:\+\efI01iR_"8C<.Goo=ZpV[i;Xq;Q8HkjJFV,&=m7QLFF@!jiZYqaT,E!BtCgfRT*/Hq7t6l'5UE,(W_B2ep(:W'GZ<:CQW_sP*IrWF]Ol5jl^4PIRF^_EZ)f0.$AA7L<1\J_O4MtW81$b>(m$ZI>>?BRRfNoaLZ*_*M?Pp6_M9]HDBc=@\o^9;+nG_m_^K.8aF'V<ZDLqI`=L';+.$q5qHs:bF+`lbp_NI2^1i1H&Q&;i[I%7QX=>j*/.`&iHHi=kc0+[@ZaQMi6u.lj3^q#d\(e2VF(\u^ScU6V<D'LSs/gJ.ML)o=FabM$rIjh$D[*tb7S#B-ZKYWV;sEO(En;^Ts7Dr7`(K95sRf$H(9\?mVk&mU*5Um^e('ek@>__GXsg%\;S6]]LQ5C/@ZG:Y]Jrh1TeJF7e(>WMEW;Gn'#nf9fu370;?"%,Ck%OKN#I?3??u/@M+;-+!d)DG6VaYHEdkj<;I^]K_.5@?U0igo/Q]=S?qm.`;"44+s8mU.7M([UUQ@,ns,rrZ@*%#/X]V4T6q6c125`^!$uh;:G`-r'9!cT*/2'Ab#YmO>2rGEWUiCT@d0JA=WH8u2Wseag-<m9>a:R\+,5l^[6at:'5pQq+4?H]<]RUc?^mO6>%NE;--nJTZZ_h/@nhFm8C5u!>?S_Ki'l73/U6upQe'<:8\<!]'eWPP$2cJ&#/8\X!L4p$.!T:e\E:hmK+(7GKs@8`0.)L)?*,\fD"]^tN:gcFr'52,e<.O;4N?:D1K?Xepi#f==O*@M`J6/HM<2+pLmDPB!Y)$l[<7S1Yc?d[Lqh*I`-UKdiO9k6*V6E*8J:l:C27/1.=MAJbp6I6neJOh6.W5dV=fG\,DMPfhJp/"3*e#KDRC)Ko+9as=S3/Vk.)^6rG]u!V6D8sdX'a1M,bs34KWtMgFY>9MpBGX)R3U#ma1im\9W1?:2Nht*3:,V@$<Qh4nqYI\I]4r]gDS/i^C3/crZ+fCqK]9UpV`?Cbo=,hDk[Y>rN]cD[N_9RAW[W#QJ^\E!'NoEP#g8<P5%DRnM$bYBt`?YraQ]@ZflW8JH["d<*[&Ygsr/BYGO\.M;VTHi1Jfbol1Ob^rs@n'75D?`fc>k.4i6I)Y]c@WghT^LkI,GSC"PG$%DR*CfsL>9.tEoHb<&;^ljQ1<WCT'n+KkX[KR$d)ilus.fsaq[fVZ7Me3Qai40ik\K@[SK%C+r-&>E'qn:Ss8"(m)9;J9qW5J8O,n/~>endstream
+endobj
+xref
+0 9
+0000000000 65535 f 
+0000000073 00000 n 
+0000000114 00000 n 
+0000000221 00000 n 
+0000000330 00000 n 
+0000000533 00000 n 
+0000000601 00000 n 
+0000000897 00000 n 
+0000000956 00000 n 
+trailer
+<<
+/ID 
+[<1c25949f9492116853fbf25e4239dab0><1c25949f9492116853fbf25e4239dab0>]
+% ReportLab generated PDF document -- digest (http://www.reportlab.com)
+
+/Info 6 0 R
+/Root 5 0 R
+/Size 9
+>>
+startxref
+3248
+%%EOF

sample/Mutual_NDA.pdf

@@ -0,0 +1,74 @@
+%PDF-1.3
+%���� ReportLab Generated PDF document http://www.reportlab.com
+1 0 obj
+<<
+/F1 2 0 R /F2 3 0 R
+>>
+endobj
+2 0 obj
+<<
+/BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font
+>>
+endobj
+3 0 obj
+<<
+/BaseFont /Times-Roman /Encoding /WinAnsiEncoding /Name /F2 /Subtype /Type1 /Type /Font
+>>
+endobj
+4 0 obj
+<<
+/Contents 8 0 R /MediaBox [ 0 0 595.2756 841.8898 ] /Parent 7 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+5 0 obj
+<<
+/PageMode /UseNone /Pages 7 0 R /Type /Catalog
+>>
+endobj
+6 0 obj
+<<
+/Author (anonymous) /CreationDate (D:20250921125754+00'00') /Creator (ReportLab PDF Library - www.reportlab.com) /Keywords () /ModDate (D:20250921125754+00'00') /Producer (ReportLab PDF Library - www.reportlab.com) 
+  /Subject (unspecified) /Title (untitled) /Trapped /False
+>>
+endobj
+7 0 obj
+<<
+/Count 1 /Kids [ 4 0 R ] /Type /Pages
+>>
+endobj
+8 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 2422
+>>
+stream
+Gat=,gN)%.&q/)-FSTK1>?^F8\]f<sUp9.m[i@CSaci9q]Kg"#6:)8h!0j7nhpVS/^:[.9S2*Y9h0o.ll%WjC$i\Dfs4MtepGMp8G"/eG>&S-KDYfCEGEf2di;K`[qYnD)_r\l!N.1`mrH*5YiKDasCq?:m/8TVo]oYtOT,uPoo/jOt?ehM;a^Ugdn'^l91gf%bidUFDRrie$GJAH+-8bb49Ti1Rnh[D_dJ=q15Nh>sdM;_?6g3jHh_UhJGF&'?V*FkR1MK>aPr#Bp-bNp33U+s2C/(9M7qPE"1YHY+oD6OMUHI#f'R"?G5;JRIUKpn;Vop*/<n><P9?pW]qfcDn=_2qXF;CIo_P56gS[BYF-r.Tf=Ja]ee*.-jfp$B;gN8R\e,.+aqhFi$3kI<3<p,9.X+Jc?C[=?4D9=XqYUUZmU-O[nlNo`B):g>VWsUslm60JeM_iqK3l;ah@0q+"`XW[)A'b2bBWqt>@\eNN+pbe-0UEMs/iYt&Of:JYqOi8@:R`9$Z#`kZE,S'I?)fXnkB#/ko4gVK8u`kDQ9EfEcLu3PejVN4b>)jo-$CM'jm;V7I]9F=`h+_`i(RuD#4+?G!63;,i;EZu3b(.)Q-K$t>cQfEA/?HQ9K=hte]sWd67m;(Q6\qE@&C5u]s=Qf+K!S3E3=ud44")OZ9d&r$,$<HfH):c3IS&I"h<;\nkg-E"AsOpH"X?E4G0tt.cVHdF;S47_?.3u<Q!Y<=;b#fE,J@ZMs6G:\dHaoX/,L_!p!"-hN7hV01=Qa67SuC08m4b+LL4'f*ElP;PfV0[Q.?.C1iM#8&gg\-/(Ep=sZ`jKi1Ea^t[4BTi;bD9W:S.#<mYL)nq[$0mFkBPXR$j\dT8o-:.OSfSRhLd(9Q'.Bo/T"N_R@HE3eji"=a64@h<pEX9n.F'r`6'Y%KZhO-]P!&%ebSrt3a&C1EH4YUo\X'tDGPN>?O[GbqC>Z[\a(Q&p&2[</CP\&1.+&ub'1<-"i;mDZE<s"5-cD.jI9)a0cpTPS%f'=HcnMTb)YcNpNSaHFd<E2o=iPL?.":26lOs+H<:`YfRe>D0<^)M'lC1K3>YL_484fgpG9Y&p/L%sQJrY#aUK"2?74\GG@DXXjCOD99R\<dkYq?cG!*/)7ehp;YNWYl)/*1Onq#,g0o*-/d$jo@"E.\TM``eL'o3cj&CglAWPaI(Tc!qsC#kL#qTF0<$;c`ifnqN%VMVS<*rA%2b"0Fm8KhLbN#i=3NsPElIc?;'pX-1@_<:6qMuI]W9_ZYS>*.fps7S1.h1c`smp;H\(:4B2gJo9`XM"U*Z`JJ5EI>p1Yf3HKk7q#o^M#VaW?r#S<,!Z$S?Y8^##$S"+IU*"QY`a*-N.*_ut>YiAV8A2N8ZU8n?4t&"/1HLO=<N')$<T%bfFeKksn@bIB&gP^sR]M5TG1$kDMZafcml6&F0P,'(\N6Z0b><:I.D[#Ara4Ku):n*D>?s^g&IH0<!5o=gQHkl[D0nYW`13$tJI=kLR^85[%0?ZH"+ARnKPl@^:^e9<.Tm:JV6uN0EufTpT<3[3!Ai$/oP;R7g?m7HV1+0A$`AZ.^51U\0n18LKfI$L[_0u8_3a$2[WT<<WmPIDG>Q$#GgUU'EDA\g:-L=p'dh=RR+fE]qIRBGc8!$C1Grh\[V]e4p+'^VP)#D1Nu$o1ED_Xm)e#;;iB*p*dm.E_h57C4-\XPE'-.A3.hiP/V"gl_&Le;Q-J7"JZf%DPdVFoNn:&.ip$sh&^WM3qGNj7U(_.9PD"EIHIg?f"$KY(fE#9l;47,G3)qDa#_aN+X'6!LD73[H@e-t\F68m`FOr2*NUU0X8X\Jo8nYk,:^UP/#+]GX6/$g:P,R)sH:ch7oo3$;XkAO9k(H9I67=(G0#:@*Nq++L_k!Q,W-Q-L)/?-TaR&Tf%*X[b1)mc$)YuSlnS]HQs@>ALCbe?Q3QC55/C86^6WUUO'e>/qL1k!tP&#]iJ%[u&]o-d^rkt7s_H#IN@-he9kR2og<=UjWabb6fWY,A]k9:eVqhR&e*mS7K:+C7h-j8GUlEfht<dbXiE(&d<(RBCGUlS5%cp'oOOc9Tt(R&[#Z@#h;*jU_$T)4N/@/4%n,+]iTP7s7<DTXWaJE9.=+'UH39I\a`;n>]J"Mjgk9C8;5iVZDNk$3gOdFK4?"keS5cYI9L$e1EAh_t:F`=:K@n)*DDY%mKhjVmi2>n,n$F7K*XX.3E&8]n6mGV$F-V(JjIc^L0fh/.M'o=0tR"q>BtdY1g'g*2(fqp)L%\;JS;8GYE*u@342]T(gP:q)<SP0(iBYQ:oJl/ZL5E3@Me?%T=+tfQ4Q>i*)rJfE,Lq'cM3&gP-\4#qA5)Td]GO1:UhIM.-lf\tC_@HBG:0o];-ERG?[p7\PK>#QN_`(<E,pqmAX8F=*1D)g_+*"HqmJ~>endstream
+endobj
+xref
+0 9
+0000000000 65535 f 
+0000000073 00000 n 
+0000000114 00000 n 
+0000000221 00000 n 
+0000000330 00000 n 
+0000000533 00000 n 
+0000000601 00000 n 
+0000000897 00000 n 
+0000000956 00000 n 
+trailer
+<<
+/ID 
+[<a017cb121e02b55d6f3b7490268d6807><a017cb121e02b55d6f3b7490268d6807>]
+% ReportLab generated PDF document -- digest (http://www.reportlab.com)
+
+/Info 6 0 R
+/Root 5 0 R
+/Size 9
+>>
+startxref
+3469
+%%EOF

sample/Residential_Lease_Agreement.pdf

@@ -0,0 +1,74 @@
+%PDF-1.3
+%���� ReportLab Generated PDF document http://www.reportlab.com
+1 0 obj
+<<
+/F1 2 0 R /F2 3 0 R
+>>
+endobj
+2 0 obj
+<<
+/BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font
+>>
+endobj
+3 0 obj
+<<
+/BaseFont /Times-Roman /Encoding /WinAnsiEncoding /Name /F2 /Subtype /Type1 /Type /Font
+>>
+endobj
+4 0 obj
+<<
+/Contents 8 0 R /MediaBox [ 0 0 595.2756 841.8898 ] /Parent 7 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+5 0 obj
+<<
+/PageMode /UseNone /Pages 7 0 R /Type /Catalog
+>>
+endobj
+6 0 obj
+<<
+/Author (anonymous) /CreationDate (D:20250921125755+00'00') /Creator (ReportLab PDF Library - www.reportlab.com) /Keywords () /ModDate (D:20250921125755+00'00') /Producer (ReportLab PDF Library - www.reportlab.com) 
+  /Subject (unspecified) /Title (untitled) /Trapped /False
+>>
+endobj
+7 0 obj
+<<
+/Count 1 /Kids [ 4 0 R ] /Type /Pages
+>>
+endobj
+8 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 1712
+>>
+stream
+Gas1_d;I\u&:EqJ/GdfeOOM`oRl9JoRjb69`sHSDMP/2r/ZiOV-q[!-EW,Vj9A-Tc["lGJ9p%M!Za3;akRIM\nUFCJIEq)\b7T_/-stEd?1O$H$)RG*XH0eohk)6nB.%=:qXF$'1-.K[UV+I;',lkiTsr.#*'0OboB[Zlk1OuGPB8o_Bm_d2dq^MJVYg\\Iuo&<nh8[drb"$857/p<QP3Dk?[dK!g/Fh?-=ohZ)q].%g4l&\Vg:q?3F+8J#l9DB63X7*L3AYKEb`F!c"q+2hU`5;0.0j[[4<Q;m#WU=ng%kKCqt?PaEj)tm4>rrk39O.E3tNi.jP9p,:Rt+C0d'k>N<6AZ<X[gS6d=b,lVXcQ4BPu41_oUD3C$"(I,]*!o[8m8)R_LQu>hd70F^+[2gT@P#.aqY4me85_sa+&EHp8(rW@&[KTTHl.c%-#QBE*i^:Z;mBf16Q>$Zu'J8uROGK#es#$8nOfh1nSUa(]1FmrKXYlbC^90'"GXu%lQP,H2$E0jJLU@UBO@?9]SjP\[',Y>MJ"h47IQ,I@Hb)adDrt2O^p"KbbI:?<%c;*D#pH+49kP$NW`pD`W,*\o&Up^M%ohNk2R+`4h8t<S&PI"Vhh1.n'g*;7dp^]g_\3"(.O-eQdk8]l+5r8lEPX&g.E.diD%>iIe9]dT:>*M*[G?I(d7[K=7sJ^`Ri5kC[B"hcB'!Ki!ko'hH^E+84T.Z,kN@&uj6?C<cjW"FD<W4SetXS+2@Rh5&a%RhP1RK`=pKod^B$1$3>cJ,@%1k7/_E3_>nh-#Iba5>KC&fKFV'?WGn1t`A2iM1Nd_KAJlF^qAl`GHmDU><L@5ee*bWqaSUqoMcI1He2Oj`d')sXE6ZDQ;'kgb)3F,c.PhV#I9._GLBRn+Q<>1=q%siLt_&a$I&LJgI+G'st)Xdf`TFL1d(P]C#2[\Cdr`O1RBJ#4VXg<sKZ"Q?m?mNpV4daWX?/n!<NA7.I3(`*`cVUIeKZ<V/Fu?u$)X-V\,gGpsHM&5JUe]'ZrFc_Q;1RS)_dOse1i;''Sh/<+S<g8'(]a^T.J_Tr4kP(Wd:.GRELoUmI;RL9!_Z)KP9Ghp#s/'+HoK0DcI_;LFuY5_jfV`0^M6K!`Jq3*D5qIQGB:RI_F]rtO6,)TdGoYg.SPGC9#VoMJpf_</)X?eq?RC@_53HB:J)QPeQ"oh=hp+'V?2Lj0tnF#/[^#&"V_X87TWat+fao(:Sd0X%3q!0&&q@5-Z8N?No8Yp0N@7G6=7a)ZQ,SD"sb[hO7U!]Ap-u52qPO&9Ub#Y2]#^)A$Xl8kic[P`=:42CI5%8N%Q<hQ!0h\(>?k%o)YPL/%^/-C99+1UI:Fim'/F1L4;*\#N>1lTgB\8QNf3triip$-p-;+(skGKm@25:c/7c%:Mu-6Y%gn0\5#NE(A&D\h>4J$k?&(8n;n+";1.0Fl.rIJ;n6gol&-tEi5UlhQc2.LL=$n#_+YFK#_(,e>/hPciB3gIH1p[r-D;r`3(;`QAGl*6NH:>-Rbl@td:IXV$i.MNqoSFX$!<43O5pFbp2<\jE/46e8"sJEY!3.D:manQ0'(GM!V)rjK6>-)@d6ECQpG=6`)g5(ken6''RVl5qR#=($"hch79$7Y"O]`47b0`T`kS5ooBuRtLM'Z/VXuahVs/S(?b`DCo]G_;X6%R6H7O]Y(&j'E4#Yt[Z/<5ip+U_EVAK75c\Sfd~>endstream
+endobj
+xref
+0 9
+0000000000 65535 f 
+0000000073 00000 n 
+0000000114 00000 n 
+0000000221 00000 n 
+0000000330 00000 n 
+0000000533 00000 n 
+0000000601 00000 n 
+0000000897 00000 n 
+0000000956 00000 n 
+trailer
+<<
+/ID 
+[<9d81045db1dd6bf8c79b9710d4d13b3d><9d81045db1dd6bf8c79b9710d4d13b3d>]
+% ReportLab generated PDF document -- digest (http://www.reportlab.com)
+
+/Info 6 0 R
+/Root 5 0 R
+/Size 9
+>>
+startxref
+2759
+%%EOF

setup.py

@@ -0,0 +1,102 @@
+#!/usr/bin/env python3
+"""
+Lega.AI Setup Script
+===================
+Interactive setup script to help configure your Lega.AI environment.
+"""
+
+import os
+import sys
+from pathlib import Path
+
+
+def main():
+    print("🚀 Welcome to Lega.AI Setup!")
+    print("=" * 50)
+    print()
+
+    # Check if .env exists
+    env_file = Path(".env")
+    if env_file.exists():
+        print("📋 Found existing .env file")
+        overwrite = input("Do you want to update it? (y/N): ").lower().strip()
+        if overwrite != "y":
+            print("Setup cancelled.")
+            return
+    else:
+        print("📋 Creating new .env file...")
+
+    # Copy from template
+    template_file = Path(".env.example")
+    if not template_file.exists():
+        print("❌ .env.example template not found!")
+        return
+
+    # Get API key from user
+    print()
+    print("🔑 Google AI API Key Setup")
+    print("-" * 30)
+    print("Get your API key from: https://makersuite.google.com/app/apikey")
+    print()
+
+    api_key = input("Enter your Google AI API key: ").strip()
+
+    if not api_key:
+        print("❌ No API key provided. You can add it later to the .env file.")
+        api_key = "your_google_ai_api_key_here"
+    else:
+        print("✅ API key received")
+
+    # Read template and replace API key
+    with open(template_file, "r") as f:
+        content = f.read()
+
+    # Replace the API key placeholder
+    content = content.replace(
+        "GOOGLE_API_KEY=your-google-api-key-here", f"GOOGLE_API_KEY={api_key}"
+    )
+
+    # Write to .env
+    with open(env_file, "w") as f:
+        f.write(content)
+
+    print()
+    print("✅ Environment file created successfully!")
+    print()
+
+    # Optional configuration
+    print("⚙️  Optional Configuration")
+    print("-" * 25)
+
+    # File size limit
+    max_size = input("Maximum file size in MB (default: 10): ").strip()
+    if max_size and max_size.isdigit():
+        content = content.replace("MAX_FILE_SIZE_MB=10", f"MAX_FILE_SIZE_MB={max_size}")
+
+    # Risk sensitivity
+    print()
+    print("Risk sensitivity (1-5, where 5 is most sensitive):")
+    risk_sens = input("Enter risk sensitivity (default: 3): ").strip()
+    if risk_sens and risk_sens.isdigit() and 1 <= int(risk_sens) <= 5:
+        content = content.replace("RISK_SENSITIVITY=3", f"RISK_SENSITIVITY={risk_sens}")
+
+    # Write updated content
+    with open(env_file, "w") as f:
+        f.write(content)
+
+    print()
+    print("🎉 Setup Complete!")
+    print("=" * 20)
+    print()
+    print("Next steps:")
+    print(
+        "1. Install dependencies: uv add streamlit 'langchain[google-genai]' langchain-google-genai langchain-chroma"
+    )
+    print("2. Run the application: streamlit run main.py")
+    print("3. Open your browser to: http://localhost:8501")
+    print()
+    print("Need help? Check the README.md file for detailed instructions.")
+
+
+if __name__ == "__main__":
+    main()

src/__init__.py

@@ -0,0 +1,3 @@
+# Create __init__.py files to make directories proper Python packages
+
+# src/__init__.py

src/models/__init__.py

@@ -0,0 +1 @@
+# src/models/__init__.py

src/models/document.py

@@ -0,0 +1,88 @@
+from pydantic import BaseModel, Field
+from typing import List, Optional, Dict, Any
+from datetime import datetime
+from enum import Enum
+
+
+class DocumentType(str, Enum):
+    RENTAL = "rental"
+    LOAN = "loan"
+    EMPLOYMENT = "employment"
+    SERVICE = "service"
+    NDA = "nda"
+    OTHER = "other"
+
+
+class RiskLevel(str, Enum):
+    LOW = "low"
+    MEDIUM = "medium"
+    HIGH = "high"
+    CRITICAL = "critical"
+
+
+class RiskCategory(str, Enum):
+    FINANCIAL = "financial"
+    COMMITMENT = "commitment"
+    RIGHTS = "rights"
+    STANDARD = "standard"
+
+
+class ClausePosition(BaseModel):
+    start_index: int
+    end_index: int
+    page_number: Optional[int] = None
+
+
+class RiskFactor(BaseModel):
+    id: str
+    clause_text: str
+    category: RiskCategory
+    severity: RiskLevel
+    explanation: str
+    suggestion: Optional[str] = None
+    position: Optional[ClausePosition] = None
+
+
+class DocumentAnalysis(BaseModel):
+    document_id: str
+    document_type: DocumentType
+    risk_score: int = Field(ge=0, le=100)
+    summary: str
+    simplified_text: str
+    risk_factors: List[RiskFactor] = []
+    key_dates: List[Dict[str, Any]] = []
+    financial_terms: Dict[str, Any] = {}
+    created_at: datetime = Field(default_factory=datetime.now)
+
+
+class Document(BaseModel):
+    id: str
+    filename: str
+    file_path: str
+    document_type: Optional[DocumentType] = None
+    file_size: int
+    upload_timestamp: datetime = Field(default_factory=datetime.now)
+    analysis: Optional[DocumentAnalysis] = None
+    processed: bool = False
+
+
+class QASession(BaseModel):
+    id: str
+    document_id: str
+    question: str
+    answer: str
+    timestamp: datetime = Field(default_factory=datetime.now)
+    confidence_score: Optional[float] = None
+
+
+class SimplificationRequest(BaseModel):
+    text: str
+    context: Optional[str] = None
+    document_type: Optional[DocumentType] = None
+
+
+class SimplificationResponse(BaseModel):
+    original_text: str
+    simplified_text: str
+    key_points: List[str] = []
+    jargon_definitions: Dict[str, str] = {}

src/pages/__init__.py

@@ -0,0 +1 @@
+# src/pages/__init__.py

src/pages/analysis.py

@@ -0,0 +1,978 @@
+import streamlit as st
+import plotly.graph_objects as go
+import plotly.express as px
+from typing import Dict, Any
+import time
+
+from ..utils.helpers import get_risk_color, extract_financial_terms, extract_key_dates
+
+
+def create_advanced_highlighting(
+    text: str, risk_factors: list, jargon_definitions: dict
+) -> str:
+    """Create advanced highlighting with hover tooltips for clauses and jargon."""
+    import re
+    
+    highlighted_text = text
+    processed_positions = []  # Track processed positions to avoid overlaps
+    
+    # First, collect all risk factors and their positions
+    risk_replacements = []
+    for i, factor in enumerate(risk_factors):
+        clause_text = factor.get("clause_text", "")
+        if not clause_text:
+            continue
+            
+        # Clean and limit clause text
+        clause_text = clause_text.strip()[:150]  # Increase limit slightly
+        
+        # Find the position in text
+        start_pos = highlighted_text.find(clause_text)
+        if start_pos != -1:
+            end_pos = start_pos + len(clause_text)
+            
+            severity = factor.get("severity", "low")
+            explanation = factor.get("explanation", "")[:200]  # Limit explanation
+            suggestion = factor.get("suggestion", "")[:200]  # Limit suggestion
+            
+            # Clean the text content for HTML (escape quotes and special chars)
+            clean_explanation = explanation.replace('"', "'").replace('<', '&lt;').replace('>', '&gt;')
+            clean_suggestion = suggestion.replace('"', "'").replace('<', '&lt;').replace('>', '&gt;')
+            
+            tooltip_content = f"⚠️ Risk: {severity.upper()}<br>📝 {clean_explanation}"
+            if clean_suggestion:
+                tooltip_content += f"<br>💡 Suggestion: {clean_suggestion}"
+            
+            risk_replacements.append({
+                'start': start_pos,
+                'end': end_pos,
+                'original': clause_text,
+                'replacement': f'<span class="tooltip risk-{severity}" title="{tooltip_content}">{clause_text}</span>',
+                'type': 'risk'
+            })
+    
+    # Sort by position (reverse order to maintain positions when replacing)
+    risk_replacements.sort(key=lambda x: x['start'], reverse=True)
+    
+    # Apply risk replacements
+    for replacement in risk_replacements:
+        start, end = replacement['start'], replacement['end']
+        highlighted_text = (
+            highlighted_text[:start] + 
+            replacement['replacement'] + 
+            highlighted_text[end:]
+        )
+        processed_positions.extend(range(start, end))
+    
+    # Then highlight jargon terms (but avoid areas already processed)
+    jargon_replacements = []
+    for term, definition in jargon_definitions.items():
+        if len(term) < 3:  # Skip very short terms
+            continue
+            
+        # Clean definition for HTML
+        clean_definition = definition.replace('"', "'").replace('<', '&lt;').replace('>', '&gt;')[:150]
+        
+        # Find all occurrences of the term (case-insensitive)
+        pattern = re.compile(r'\b' + re.escape(term) + r'\b', re.IGNORECASE)
+        
+        for match in pattern.finditer(highlighted_text):
+            start_pos, end_pos = match.span()
+            
+            # Check if this position overlaps with existing highlights
+            if any(pos in processed_positions for pos in range(start_pos, end_pos)):
+                continue
+                
+            # Check if we're inside an HTML tag
+            before_text = highlighted_text[:start_pos]
+            if before_text.count('<span') > before_text.count('</span>'):
+                continue  # We're inside a span, skip
+                
+            jargon_replacements.append({
+                'start': start_pos,
+                'end': end_pos,
+                'original': match.group(),
+                'replacement': f'<span class="tooltip jargon-term" title="📚 {term}: {clean_definition}">{match.group()}</span>',
+                'type': 'jargon'
+            })
+    
+    # Sort jargon replacements by position (reverse order)
+    jargon_replacements.sort(key=lambda x: x['start'], reverse=True)
+    
+    # Apply jargon replacements (limit to 5 to avoid clutter)
+    for replacement in jargon_replacements[:5]:
+        start, end = replacement['start'], replacement['end']
+        highlighted_text = (
+            highlighted_text[:start] + 
+            replacement['replacement'] + 
+            highlighted_text[end:]
+        )
+    
+    return highlighted_text
+
+
+def show_analysis_interface():
+    """Display the document analysis interface."""
+
+    if not st.session_state.get("current_document"):
+        st.info("📊 **Document Analysis Page**")
+        st.markdown("### No document selected for analysis")
+        st.markdown("""
+        To view analysis results, you need to:
+        1. **Upload a new document** for instant analysis, or
+        2. **Check your library** for previously analyzed documents
+        """)
+        
+        col1, col2, col3 = st.columns(3)
+        
+        with col1:
+            if st.button("📄 Upload Document", type="primary", use_container_width=True):
+                st.session_state.page = "📄 Upload"
+                st.rerun()
+        
+        with col2:
+            if st.button("📚 View Library", use_container_width=True):
+                st.session_state.page = "� Library"
+                st.rerun()
+                
+        with col3:
+            if st.button("🏠 Go Home", use_container_width=True):
+                st.session_state.page = "🏠 Home"
+                st.rerun()
+        
+        # Show recently analyzed documents if available
+        if st.session_state.get("documents_library"):
+            st.markdown("---")
+            st.markdown("### 📋 Recently Analyzed Documents")
+            st.markdown("Click on any document below to view its analysis:")
+            
+            for doc in st.session_state.documents_library[-3:]:  # Show last 3
+                col1, col2 = st.columns([3, 1])
+                with col1:
+                    st.markdown(f"**{doc.get('filename', 'Unknown')}** - {doc.get('document_type', 'Unknown').title()}")
+                with col2:
+                    if st.button(f"View Analysis", key=f"view_{doc.get('id')}", use_container_width=True):
+                        # Load this document for analysis
+                        st.session_state.current_document = doc
+                        st.rerun()
+        
+        return
+
+    doc = st.session_state.current_document
+
+    # Header
+    st.header("📊 Document Analysis")
+    st.markdown(
+        f"**File:** {doc.get('filename', 'Unknown')} | **Type:** {doc.get('document_type', 'Unknown').title()}"
+    )
+
+    # If it's a sample document, process it first
+    if doc.get("is_sample") and not doc.get("processed"):
+        process_sample_document(doc)
+        return
+
+    # Risk Score Dashboard
+    show_risk_dashboard(doc)
+
+    # Document Content Analysis
+    col1, col2 = st.columns([1, 1])
+
+    with col1:
+        show_original_document(doc)
+
+    with col2:
+        show_simplified_version(doc)
+
+    # Additional Analysis Sections
+    st.markdown("---")
+
+    # Tabs for different analysis views
+    tab1, tab2, tab3, tab4, tab5 = st.tabs(
+        [
+            "📋 Summary",
+            "⚠️ Risk Factors",
+            "📅 Key Dates",
+            "💰 Financial Terms",
+            "📊 Market Comparison",
+        ]
+    )
+
+    with tab1:
+        show_document_summary(doc)
+
+    with tab2:
+        show_risk_factors(doc)
+
+    with tab3:
+        show_key_dates(doc)
+
+    with tab4:
+        show_financial_terms(doc)
+
+    with tab5:
+        show_market_comparison(doc)
+
+    # Action buttons
+    st.markdown("---")
+    col1, col2, col3 = st.columns(3)
+
+    with col1:
+        if st.button("💬 Ask Questions", use_container_width=True):
+            st.session_state.page = "💬 Q&A"
+            st.rerun()
+
+    with col2:
+        if st.button("📥 Export Report", use_container_width=True):
+            export_report(doc)
+
+    with col3:
+        if st.button("📄 Analyze New Document", use_container_width=True):
+            st.session_state.current_document = None
+            st.session_state.page = "📄 Upload"
+            st.rerun()
+
+
+def process_sample_document(doc):
+    """Process a sample document with simulated AI analysis."""
+    st.info("🤖 Processing sample document with AI analysis...")
+
+    progress_bar = st.progress(0)
+    status_text = st.empty()
+
+    # Simulate processing steps
+    steps = [
+        ("📄 Extracting text...", 20),
+        ("🔍 Detecting document type...", 40),
+        ("⚠️ Analyzing risks...", 60),
+        ("💬 Simplifying language...", 80),
+        ("📋 Generating summary...", 100),
+    ]
+
+    for step_text, progress in steps:
+        status_text.text(step_text)
+        progress_bar.progress(progress)
+        time.sleep(0.5)
+
+    # Generate mock analysis results
+    doc_type = doc.get("document_type", "other")
+
+    # Mock risk factors based on document type
+    risk_factors = generate_mock_risk_factors(doc_type)
+    simplified_text = generate_mock_simplified_text(
+        doc.get("original_text", ""), doc_type
+    )
+    summary = generate_mock_summary(doc_type)
+
+    # Update document with analysis
+    doc.update(
+        {
+            "risk_data": {
+                "risk_factors": risk_factors,
+                "overall_assessment": f"This {doc_type} document contains several high-risk clauses.",
+            },
+            "simplified_text": simplified_text,
+            "summary": summary,
+            "key_points": [
+                f"Key point 1 for {doc_type}",
+                f"Key point 2 for {doc_type}",
+                f"Key point 3 for {doc_type}",
+            ],
+            "jargon_definitions": {
+                "Liability": "Legal responsibility for damages",
+                "Arbitration": "Dispute resolution outside of court",
+            },
+            "processed": True,
+            "analysis_timestamp": time.time(),
+        }
+    )
+
+    st.session_state.current_document = doc
+
+    progress_bar.empty()
+    status_text.empty()
+    st.success("✅ Analysis complete!")
+    time.sleep(1)
+    st.rerun()
+
+
+def show_risk_dashboard(doc):
+    """Display the risk assessment dashboard."""
+    risk_data = doc.get("risk_data", {})
+    risk_factors = risk_data.get("risk_factors", [])
+
+    # Calculate risk score
+    risk_score = min(len(risk_factors) * 15, 100)
+
+    # Risk score gauge
+    col1, col2, col3 = st.columns([2, 1, 1])
+
+    with col1:
+        # Create gauge chart
+        fig = go.Figure(
+            go.Indicator(
+                mode="gauge+number+delta",
+                value=risk_score,
+                domain={"x": [0, 1], "y": [0, 1]},
+                title={"text": "Risk Score"},
+                delta={"reference": 50},
+                gauge={
+                    "axis": {"range": [None, 100]},
+                    "bar": {"color": get_risk_color(risk_score)},
+                    "steps": [
+                        {"range": [0, 25], "color": "lightgray"},
+                        {"range": [25, 50], "color": "gray"},
+                        {"range": [50, 75], "color": "lightcoral"},
+                        {"range": [75, 100], "color": "red"},
+                    ],
+                    "threshold": {
+                        "line": {"color": "red", "width": 4},
+                        "thickness": 0.75,
+                        "value": 90,
+                    },
+                },
+            )
+        )
+
+        fig.update_layout(height=300)
+        st.plotly_chart(fig, use_container_width=True)
+
+    with col2:
+        st.metric(
+            label="Risk Factors Found",
+            value=len(risk_factors),
+            delta=f"vs average: +{max(0, len(risk_factors) - 3)}",
+        )
+
+    with col3:
+        risk_level = (
+            "Low"
+            if risk_score < 25
+            else (
+                "Medium"
+                if risk_score < 50
+                else "High" if risk_score < 75 else "Critical"
+            )
+        )
+        st.metric(
+            label="Risk Level",
+            value=risk_level,
+            delta_color="inverse" if risk_score > 50 else "normal",
+        )
+
+    # Risk assessment summary
+    if risk_data.get("overall_assessment"):
+        st.info(f"**Assessment:** {risk_data['overall_assessment']}")
+
+
+def show_original_document(doc):
+    """Display the original document with advanced highlighting and hover definitions."""
+    st.subheader("📄 Original Document")
+
+    original_text = doc.get("original_text", "")
+    risk_factors = doc.get("risk_data", {}).get("risk_factors", [])
+    jargon_definitions = doc.get("jargon_definitions", {})
+
+    # Advanced highlighting with hover tooltips
+    highlighted_text = create_advanced_highlighting(
+        original_text, risk_factors, jargon_definitions
+    )
+
+    # Custom CSS for hover tooltips with responsive theming
+    st.markdown(
+        """
+    <style>
+    .tooltip {
+        position: relative;
+        display: inline;
+        cursor: help;
+        border-radius: 4px;
+        padding: 2px 4px;
+        margin: 0 1px;
+    }
+    
+    /* Risk highlighting with theme-aware backgrounds */
+    .risk-critical { 
+        background-color: rgba(255, 68, 68, 0.2); 
+        border-left: 4px solid #ff4444; 
+        padding: 4px 8px;
+        border-radius: 4px;
+        cursor: help;
+    }
+    .risk-high { 
+        background-color: rgba(255, 136, 0, 0.2); 
+        border-left: 4px solid #ff8800; 
+        padding: 4px 8px;
+        border-radius: 4px;
+        cursor: help;
+    }
+    .risk-medium { 
+        background-color: rgba(255, 204, 0, 0.2); 
+        border-left: 4px solid #ffcc00; 
+        padding: 4px 8px;
+        border-radius: 4px;
+        cursor: help;
+    }
+    .risk-low { 
+        background-color: rgba(68, 170, 68, 0.2); 
+        border-left: 4px solid #44aa44; 
+        padding: 4px 8px;
+        border-radius: 4px;
+        cursor: help;
+    }
+    
+    /* Jargon term highlighting */
+    .jargon-term { 
+        background-color: rgba(46, 134, 171, 0.2); 
+        text-decoration: underline dotted #2e86ab;
+        padding: 2px 4px;
+        border-radius: 3px;
+        cursor: help;
+    }
+    
+    /* Enhanced tooltips */
+    .tooltip:hover {
+        opacity: 0.8;
+    }
+    </style>
+    """,
+        unsafe_allow_html=True,
+    )
+
+    st.markdown(highlighted_text, unsafe_allow_html=True)
+
+    # Scroll area for long documents
+    if len(original_text) > 1000:
+        with st.expander("View Full Document"):
+            st.text_area("Full Text", original_text, height=400, disabled=True)
+
+
+def show_simplified_version(doc):
+    """Display the simplified version of the document."""
+    st.subheader("💬 Simplified Version")
+
+    simplified_text = doc.get("simplified_text", "Processing...")
+    st.markdown(simplified_text)
+
+    # Key points
+    key_points = doc.get("key_points", [])
+    if key_points:
+        st.markdown("**Key Points:**")
+        for point in key_points:
+            st.markdown(f"• {point}")
+
+    # Jargon definitions
+    jargon_definitions = doc.get("jargon_definitions", {})
+    if jargon_definitions:
+        st.markdown("**Legal Terms Explained:**")
+        for term, definition in jargon_definitions.items():
+            st.markdown(f"**{term}:** {definition}")
+
+
+def show_document_summary(doc):
+    """Display document summary."""
+    summary = doc.get("summary", "Generating summary...")
+    st.markdown(summary)
+
+    # Document metadata
+    st.markdown("### 📊 Document Information")
+    col1, col2 = st.columns(2)
+
+    with col1:
+        st.markdown(f"**Type:** {doc.get('document_type', 'Unknown').title()}")
+        st.markdown(f"**Filename:** {doc.get('filename', 'Unknown')}")
+
+    with col2:
+        if doc.get("file_size"):
+            from ..utils.helpers import format_file_size
+
+            st.markdown(f"**Size:** {format_file_size(doc['file_size'])}")
+
+        if doc.get("analysis_timestamp"):
+            import datetime
+
+            analysis_time = datetime.datetime.fromtimestamp(doc["analysis_timestamp"])
+            st.markdown(f"**Analyzed:** {analysis_time.strftime('%Y-%m-%d %H:%M')}")
+
+
+def show_risk_factors(doc):
+    """Display detailed risk factors."""
+    risk_factors = doc.get("risk_data", {}).get("risk_factors", [])
+
+    if not risk_factors:
+        st.info("No significant risk factors identified in this document.")
+        return
+
+    for i, factor in enumerate(risk_factors):
+        severity = factor.get("severity", "low")
+
+        # Color coding based on severity
+        if severity == "critical":
+            st.error(f"🚨 **Critical Risk #{i+1}**")
+        elif severity == "high":
+            st.warning(f"⚠️ **High Risk #{i+1}**")
+        elif severity == "medium":
+            st.info(f"🟡 **Medium Risk #{i+1}**")
+        else:
+            st.success(f"🟢 **Low Risk #{i+1}**")
+
+        st.markdown(f"**Clause:** {factor.get('clause_text', 'N/A')}")
+        st.markdown(f"**Category:** {factor.get('category', 'N/A').title()}")
+        st.markdown(f"**Explanation:** {factor.get('explanation', 'N/A')}")
+
+        if factor.get("suggestion"):
+            st.markdown(f"**Suggestion:** {factor['suggestion']}")
+
+        st.markdown("---")
+
+
+def show_key_dates(doc):
+    """Display extracted key dates with timeline visualization."""
+    original_text = doc.get("original_text", "")
+    dates = extract_key_dates(original_text)
+
+    if not dates:
+        st.info("No specific dates found in this document.")
+        return
+
+    # Enhanced date analysis with timeline
+    col1, col2 = st.columns([1, 1])
+
+    with col1:
+        st.markdown("**Important Dates Found:**")
+        for date_info in dates:
+            st.markdown(f"• **{date_info['date']}** - Context: {date_info['context']}")
+
+    with col2:
+        st.markdown("**Timeline & Obligations:**")
+
+        # Mock timeline data based on document type
+        doc_type = doc.get("document_type", "other")
+
+        if doc_type == "rental":
+            timeline_items = [
+                {
+                    "date": "1st of every month",
+                    "event": "Rent Payment Due",
+                    "type": "recurring",
+                },
+                {
+                    "date": "30 days notice",
+                    "event": "Termination Notice Required",
+                    "type": "condition",
+                },
+                {
+                    "date": "End of lease",
+                    "event": "Security Deposit Return",
+                    "type": "deadline",
+                },
+            ]
+        elif doc_type == "loan":
+            timeline_items = [
+                {
+                    "date": "15th of every month",
+                    "event": "EMI Payment Due",
+                    "type": "recurring",
+                },
+                {
+                    "date": "7 days after due",
+                    "event": "Late Fee Applicable",
+                    "type": "penalty",
+                },
+                {"date": "24 months", "event": "Loan Maturity", "type": "deadline"},
+            ]
+        elif doc_type == "employment":
+            timeline_items = [
+                {
+                    "date": "Last day of month",
+                    "event": "Salary Payment",
+                    "type": "recurring",
+                },
+                {
+                    "date": "90 days",
+                    "event": "Resignation Notice Period",
+                    "type": "condition",
+                },
+                {
+                    "date": "2 years post-termination",
+                    "event": "Non-compete Expires",
+                    "type": "deadline",
+                },
+            ]
+        else:
+            timeline_items = []
+
+        for item in timeline_items:
+            if item["type"] == "recurring":
+                st.markdown(f"🔄 **{item['date']}**: {item['event']}")
+            elif item["type"] == "penalty":
+                st.markdown(f"⚠️ **{item['date']}**: {item['event']}")
+            elif item["type"] == "deadline":
+                st.markdown(f"📅 **{item['date']}**: {item['event']}")
+            else:
+                st.markdown(f"📌 **{item['date']}**: {item['event']}")
+
+    # Visual timeline chart
+    if timeline_items:
+        st.markdown("---")
+        st.markdown("**📊 Visual Timeline**")
+
+        # Create timeline visualization
+        timeline_df = []
+        for i, item in enumerate(timeline_items):
+            timeline_df.append(
+                {
+                    "Event": item["event"],
+                    "Timeline": item["date"],
+                    "Type": item["type"].title(),
+                    "Order": i,
+                }
+            )
+
+        if timeline_df:
+            import pandas as pd
+
+            df = pd.DataFrame(timeline_df)
+
+            # Color code by type
+            color_map = {
+                "Recurring": "#2e86ab",
+                "Penalty": "#ff4444",
+                "Deadline": "#ff8800",
+                "Condition": "#44aa44",
+            }
+
+            fig = px.timeline(
+                df,
+                x_start=[0] * len(df),
+                x_end=[1] * len(df),
+                y="Event",
+                color="Type",
+                color_discrete_map=color_map,
+                title="Contract Timeline & Obligations",
+            )
+            st.plotly_chart(fig, use_container_width=True)
+
+
+def show_financial_terms(doc):
+    """Display extracted financial terms."""
+    original_text = doc.get("original_text", "")
+    financial_terms = extract_financial_terms(original_text)
+
+    if not financial_terms:
+        st.info("No financial terms identified in this document.")
+        return
+
+    col1, col2 = st.columns(2)
+
+    with col1:
+        if "amounts" in financial_terms:
+            st.markdown("**Monetary Amounts:**")
+            for amount in financial_terms["amounts"]:
+                st.markdown(f"• {amount}")
+
+    with col2:
+        if "percentages" in financial_terms:
+            st.markdown("**Percentages/Rates:**")
+            for percentage in financial_terms["percentages"]:
+                st.markdown(f"• {percentage}")
+
+    if "interest_rates" in financial_terms:
+        st.markdown("**Interest Rates:**")
+        for rate in financial_terms["interest_rates"]:
+            st.markdown(f"• {rate}")
+
+
+def export_report(doc):
+    """Export analysis report."""
+    # Create a simple text report
+    report = f"""
+LEGA.AI DOCUMENT ANALYSIS REPORT
+{'='*50}
+
+Document: {doc.get('filename', 'Unknown')}
+Type: {doc.get('document_type', 'Unknown').title()}
+Analysis Date: {time.strftime('%Y-%m-%d %H:%M:%S')}
+
+SUMMARY:
+{doc.get('summary', 'No summary available')}
+
+RISK ASSESSMENT:
+{doc.get('risk_data', {}).get('overall_assessment', 'No risk assessment available')}
+
+RISK FACTORS:
+"""
+
+    risk_factors = doc.get("risk_data", {}).get("risk_factors", [])
+    for i, factor in enumerate(risk_factors):
+        report += f"""
+{i+1}. {factor.get('severity', 'Unknown').upper()} RISK
+   Category: {factor.get('category', 'N/A').title()}
+   Clause: {factor.get('clause_text', 'N/A')}
+   Explanation: {factor.get('explanation', 'N/A')}
+"""
+
+    report += f"""
+
+SIMPLIFIED VERSION:
+{doc.get('simplified_text', 'No simplified version available')}
+
+KEY POINTS:
+"""
+
+    for point in doc.get("key_points", []):
+        report += f"• {point}\n"
+
+    report += "\n\nGenerated by Lega.AI - Making legal documents accessible"
+
+    # Clean filename - remove .pdf extension if present
+    filename = doc.get('filename', 'document')
+    if filename.endswith('.pdf'):
+        filename = filename[:-4]
+    if filename.endswith('.docx'):
+        filename = filename[:-5]
+    if filename.endswith('.txt'):
+        filename = filename[:-4]
+
+    # Offer download
+    st.download_button(
+        label="📥 Download Report",
+        data=report,
+        file_name=f"lega_ai_report_{filename}.pdf",
+        mime="application/pdf",
+    )
+
+    st.success("✅ Report prepared for download!")
+
+
+def generate_mock_risk_factors(doc_type):
+    """Generate mock risk factors for sample documents."""
+    if doc_type == "rental":
+        return [
+            {
+                "clause_text": "Late payments will incur a penalty of Rs. 1,000 per day",
+                "category": "financial",
+                "severity": "high",
+                "explanation": "Daily penalties can quickly escalate to substantial amounts",
+                "suggestion": "Negotiate a more reasonable penalty structure",
+            },
+            {
+                "clause_text": "Tenant is responsible for all repairs and maintenance",
+                "category": "financial",
+                "severity": "medium",
+                "explanation": "This places unusual burden on tenant for structural repairs",
+                "suggestion": "Clarify that structural repairs remain landlord responsibility",
+            },
+        ]
+    elif doc_type == "loan":
+        return [
+            {
+                "clause_text": "24% per annum (APR 28.5% including processing fees)",
+                "category": "financial",
+                "severity": "critical",
+                "explanation": "Interest rate is significantly above market rates",
+                "suggestion": "Shop around for better rates from other lenders",
+            },
+            {
+                "clause_text": "Lender may seize collateral immediately upon default",
+                "category": "rights",
+                "severity": "high",
+                "explanation": "No grace period or notice before asset seizure",
+                "suggestion": "Negotiate for notice period and cure opportunity",
+            },
+        ]
+    elif doc_type == "employment":
+        return [
+            {
+                "clause_text": "Employee shall not work for any competing company for 2 years",
+                "category": "commitment",
+                "severity": "high",
+                "explanation": "Non-compete period is unusually long and broad",
+                "suggestion": "Negotiate shorter period and narrower scope",
+            },
+            {
+                "clause_text": "Company may terminate employment at any time without cause",
+                "category": "rights",
+                "severity": "medium",
+                "explanation": "No job security or notice period for termination",
+                "suggestion": "Request notice period and severance terms",
+            },
+        ]
+    else:
+        return []
+
+
+def generate_mock_simplified_text(original_text, doc_type):
+    """Generate mock simplified text."""
+    if doc_type == "rental":
+        return """
+**What this rental agreement means in simple terms:**
+
+You're renting a property in Mumbai for ₹25,000 per month. Here are the key things to know:
+
+• **Payment:** You must pay rent by the 1st of each month. If you're late, you'll be charged ₹1,000 for each day you're late.
+
+• **Security deposit:** You need to pay ₹75,000 upfront as security. This money is hard to get back.
+
+• **Repairs:** You're responsible for fixing everything that breaks, even major structural problems.
+
+• **Leaving early:** If you want to leave before the lease ends, you lose your security deposit.
+
+**Watch out for:** The daily late fees and your responsibility for all repairs are unusual and costly.
+        """
+    elif doc_type == "loan":
+        return """
+**What this loan agreement means in simple terms:**
+
+You're borrowing ₹2,00,000 but will pay back ₹3,00,000 total - that's ₹1,00,000 extra in interest and fees.
+
+• **Monthly payment:** ₹12,500 every month for 2 years
+
+• **Interest rate:** 24% per year (very high - normal rates are 10-15%)
+
+• **Late fees:** ₹500 per day if you're late
+
+• **Your gold jewelry:** The lender can take it immediately if you miss payments
+
+• **Total cost:** You'll pay 50% more than you borrowed
+
+**Warning:** This is an expensive loan. The interest rate is much higher than banks typically charge.
+        """
+    elif doc_type == "employment":
+        return """
+**What this employment contract means in simple terms:**
+
+You're being hired as a Software Developer for ₹8,00,000 per year. Here's what you need to know:
+
+• **Working hours:** 45 hours per week, including weekends when needed
+
+• **Salary:** ₹66,667 per month
+
+• **If you quit:** You must give 90 days notice
+
+• **If they fire you:** They can fire you anytime without reason or notice
+
+• **After leaving:** You can't work for competing companies for 2 years
+
+• **Side work:** You can't do any other work while employed
+
+**Concerns:** The 2-year non-compete and ability to fire without notice are harsh terms.
+        """
+    else:
+        return "Document simplified version will appear here after analysis."
+
+
+def show_market_comparison(doc):
+    """Display market benchmarking and comparison data."""
+    doc_type = doc.get("document_type", "other")
+
+    st.markdown("**Market Context & Benchmarking**")
+
+    if doc_type == "rental":
+        show_rental_market_comparison(doc)
+    elif doc_type == "loan":
+        show_loan_market_comparison(doc)
+    elif doc_type == "employment":
+        show_employment_market_comparison(doc)
+    else:
+        st.info(
+            "Market comparison data available for rental, loan, and employment contracts."
+        )
+
+
+def show_rental_market_comparison(doc):
+    """Show rental market comparison."""
+    col1, col2 = st.columns(2)
+
+    with col1:
+        st.markdown("#### 🏠 Rental Market Analysis")
+        st.markdown("**Security Deposit:** ₹75,000")
+        st.success("✅ Standard: Typically 2-3 months rent")
+
+        st.markdown("**Late Penalty:** ₹1,000/day")
+        st.error("❌ Above Market: Typical penalties are ₹100-500/day")
+
+        st.markdown("**Maintenance Responsibility:** Tenant")
+        st.warning("⚠️ Unusual: Structural repairs typically landlord's responsibility")
+
+    with col2:
+        st.markdown("#### 📊 Mumbai Rental Benchmarks")
+
+        # Mock market data
+        market_data = {
+            "Average Rent (2BHK)": "₹28,000",
+            "Security Deposit Range": "₹50,000 - ₹84,000",
+            "Standard Late Fee": "₹200/day",
+            "Tenant Maintenance": "10% of agreements",
+        }
+
+        for metric, value in market_data.items():
+            st.metric(metric, value)
+
+
+def show_loan_market_comparison(doc):
+    """Show loan market comparison."""
+    col1, col2 = st.columns(2)
+
+    with col1:
+        st.markdown("#### 💰 Loan Market Analysis")
+        st.markdown("**Interest Rate:** 24% per annum")
+        st.error("❌ Well Above Market: Bank rates typically 10-15%")
+
+        st.markdown("**Processing Fee:** ₹10,000")
+        st.warning("⚠️ High: Typical processing fees 1-2% of loan amount")
+
+        st.markdown("**Total Repayment:** ₹3,00,000 for ₹2,00,000")
+        st.error("❌ Very High: 50% more than principal")
+
+    with col2:
+        st.markdown("#### 📊 Personal Loan Benchmarks")
+
+        # Create comparison chart
+        fig = px.bar(
+            x=["Your Loan", "Bank Average", "NBFC Average"],
+            y=[24, 12, 18],
+            title="Interest Rate Comparison (%)",
+            color=["red", "green", "orange"],
+        )
+        st.plotly_chart(fig, use_container_width=True)
+
+
+def show_employment_market_comparison(doc):
+    """Show employment market comparison."""
+    col1, col2 = st.columns(2)
+
+    with col1:
+        st.markdown("#### 💼 Employment Market Analysis")
+        st.markdown("**Non-compete Period:** 2 years")
+        st.error("❌ Excessive: Typical non-compete is 6-12 months")
+
+        st.markdown("**Notice Period:** 90 days")
+        st.warning("⚠️ Long: Standard notice is 30-60 days")
+
+        st.markdown("**At-will Termination:** Yes")
+        st.error("❌ Unfavorable: Most contracts provide notice period")
+
+    with col2:
+        st.markdown("#### 📊 IT Industry Standards")
+
+        standards = {
+            "Average Salary (3-5 YOE)": "₹8-12 lakhs",
+            "Standard Notice Period": "30-60 days",
+            "Typical Non-compete": "6-12 months",
+            "Weekend Work": "Occasionally, not mandatory",
+        }
+
+        for standard, value in standards.items():
+            st.metric(standard, value)
+
+
+def generate_mock_summary(doc_type):
+    """Generate mock summary."""
+    if doc_type == "rental":
+        return "This is a residential lease agreement for a property in Mumbai with rent of ₹25,000/month. The agreement contains several tenant-unfavorable terms including high daily late fees, tenant responsibility for all repairs, and forfeiture of security deposit for early termination."
+    elif doc_type == "loan":
+        return "This is a personal loan agreement for ₹2,00,000 with very high interest rates (24% APR, 28.5% effective). The loan requires gold jewelry as collateral and includes harsh default terms with immediate asset seizure rights."
+    elif doc_type == "employment":
+        return "This is an employment contract for a Software Developer position with ₹8,00,000 annual salary. The contract includes restrictive terms like a 2-year non-compete clause, at-will termination by employer, and prohibition on side work."
+    else:
+        return "Document summary will appear here after analysis."

src/pages/library.py

@@ -0,0 +1,340 @@
+import streamlit as st
+import pandas as pd
+from typing import List, Dict
+import time
+
+from ..utils.helpers import format_file_size, format_timestamp
+
+
+def show_library_interface():
+    """Display the document library interface."""
+
+    st.header("📚 Document Library")
+    st.markdown("Manage and review all your analyzed documents")
+
+    # Get documents from session state
+    documents = st.session_state.get("documents_library", [])
+
+    if not documents:
+        show_empty_library()
+        return
+
+    # Library statistics
+    show_library_stats(documents)
+
+    # Filter and search
+    show_library_filters(documents)
+
+    # Document grid
+    show_document_grid(documents)
+
+
+def show_empty_library():
+    """Show empty library state."""
+    st.markdown("---")
+
+    col1, col2, col3 = st.columns([1, 2, 1])
+
+    with col2:
+        st.markdown(
+            """
+        <div style="text-align: center; padding: 3rem;">
+            <h3>📚 Your Library is Empty</h3>
+            <p style="color: var(--text-color, #666); opacity: 0.7;">Upload and analyze documents to build your personal legal document library.</p>
+        </div>
+        """,
+            unsafe_allow_html=True,
+        )
+
+        if st.button(
+            "📄 Upload Your First Document", type="primary", use_container_width=True
+        ):
+            st.session_state.page = "📄 Upload"
+            st.rerun()
+
+    # Add sample documents section
+    st.markdown("---")
+    show_sample_documents_section()
+
+
+def show_library_stats(documents: List[Dict]):
+    """Display library statistics."""
+    # Calculate stats
+    total_docs = len(documents)
+    doc_types = {}
+    high_risk_docs = 0
+
+    for doc in documents:
+        doc_type = doc.get("document_type", "other")
+        doc_types[doc_type] = doc_types.get(doc_type, 0) + 1
+
+        if doc.get("risk_score", 0) > 60:
+            high_risk_docs += 1
+
+    # Display stats
+    col1, col2, col3, col4 = st.columns(4)
+
+    with col1:
+        st.metric(label="Total Documents", value=total_docs)
+
+    with col2:
+        most_common_type = max(doc_types, key=doc_types.get) if doc_types else "None"
+        st.metric(label="Most Common Type", value=most_common_type.title())
+
+    with col3:
+        st.metric(
+            label="High Risk Documents",
+            value=high_risk_docs,
+            delta=(
+                f"{high_risk_docs/total_docs*100:.0f}% of total"
+                if total_docs > 0
+                else "0%"
+            ),
+        )
+
+    with col4:
+        total_size = sum(doc.get("file_size", 0) for doc in documents)
+        st.metric(label="Total Storage", value=format_file_size(total_size))
+
+
+def show_library_filters(documents: List[Dict]):
+    """Display filter and search options."""
+    st.markdown("---")
+
+    col1, col2, col3 = st.columns(3)
+
+    with col1:
+        # Document type filter
+        doc_types = ["All"] + list(
+            set(doc.get("document_type", "other") for doc in documents)
+        )
+        selected_type = st.selectbox("Filter by Type", doc_types)
+
+    with col2:
+        # Risk level filter
+        risk_levels = [
+            "All",
+            "Low Risk (0-30)",
+            "Medium Risk (31-60)",
+            "High Risk (61+)",
+        ]
+        selected_risk = st.selectbox("Filter by Risk", risk_levels)
+
+    with col3:
+        # Search
+        search_term = st.text_input(
+            "Search documents", placeholder="Enter filename or content..."
+        )
+
+    # Apply filters
+    filtered_docs = documents
+
+    if selected_type != "All":
+        filtered_docs = [
+            doc for doc in filtered_docs if doc.get("document_type") == selected_type
+        ]
+
+    if selected_risk != "All":
+        if "Low Risk" in selected_risk:
+            filtered_docs = [
+                doc for doc in filtered_docs if doc.get("risk_score", 0) <= 30
+            ]
+        elif "Medium Risk" in selected_risk:
+            filtered_docs = [
+                doc for doc in filtered_docs if 31 <= doc.get("risk_score", 0) <= 60
+            ]
+        elif "High Risk" in selected_risk:
+            filtered_docs = [
+                doc for doc in filtered_docs if doc.get("risk_score", 0) > 60
+            ]
+
+    if search_term:
+        filtered_docs = [
+            doc
+            for doc in filtered_docs
+            if search_term.lower() in doc.get("filename", "").lower()
+        ]
+
+    # Store filtered docs for grid display
+    st.session_state.filtered_documents = filtered_docs
+
+
+def show_document_grid(documents: List[Dict]):
+    """Display documents in a grid layout."""
+    filtered_docs = st.session_state.get("filtered_documents", documents)
+
+    if not filtered_docs:
+        st.info("No documents match your filter criteria.")
+        return
+
+    st.markdown("---")
+    st.subheader(f"📄 Documents ({len(filtered_docs)})")
+
+    # Display documents in cards
+    for i in range(0, len(filtered_docs), 2):
+        col1, col2 = st.columns(2)
+
+        # First document
+        with col1:
+            if i < len(filtered_docs):
+                show_document_card(filtered_docs[i])
+
+        # Second document
+        with col2:
+            if i + 1 < len(filtered_docs):
+                show_document_card(filtered_docs[i + 1])
+
+
+def show_document_card(doc: Dict):
+    """Display a single document card."""
+    # Risk color
+    risk_score = doc.get("risk_score", 0)
+    if risk_score > 60:
+        risk_color = "🔴"
+        risk_label = "High Risk"
+    elif risk_score > 30:
+        risk_color = "🟠"
+        risk_label = "Medium Risk"
+    else:
+        risk_color = "🟢"
+        risk_label = "Low Risk"
+
+    # Use container for card styling
+    with st.container():
+        # Header row with filename and risk
+        col1, col2 = st.columns([3, 1])
+        with col1:
+            st.markdown(f"**📄 {doc.get('filename', 'Unknown')}**")
+        with col2:
+            st.markdown(f"{risk_color} {risk_label}")
+
+        # Document details
+        doc_type = doc.get("document_type", "other").title()
+        upload_date = doc.get("upload_date", "Unknown")
+        file_size = format_file_size(doc.get("file_size", 0))
+
+        st.markdown(f"📋 {doc_type} • 📅 {upload_date} • 💾 {file_size}")
+
+        # Add some spacing
+        st.markdown("---")
+
+    # Action buttons
+    col1, col2, col3 = st.columns(3)
+
+    with col1:
+        if st.button("📊 View", key=f"view_{doc['id']}", use_container_width=True):
+            load_document_for_analysis(doc["id"])
+
+    with col2:
+        if st.button("💬 Q&A", key=f"qa_{doc['id']}", use_container_width=True):
+            load_document_for_qa(doc["id"])
+
+    with col3:
+        if st.button("🗑️ Delete", key=f"delete_{doc['id']}", use_container_width=True):
+            delete_document(doc["id"])
+
+
+def load_document_for_analysis(doc_id: str):
+    """Load a document from library for analysis."""
+    documents = st.session_state.get("documents_library", [])
+
+    for doc in documents:
+        if doc["id"] == doc_id:
+            # Simulate loading the full document data
+            st.session_state.current_document = {
+                "id": doc["id"],
+                "filename": doc["filename"],
+                "document_type": doc["document_type"],
+                "original_text": f"Sample content for {doc['filename']}...",  # In real app, load from storage
+                "is_sample": True,  # Mark as sample for demo
+                "risk_score": doc.get("risk_score", 0),
+            }
+
+            st.session_state.page = "📊 Analysis"
+            st.rerun()
+            break
+
+
+def load_document_for_qa(doc_id: str):
+    """Load a document from library for Q&A."""
+    documents = st.session_state.get("documents_library", [])
+
+    for doc in documents:
+        if doc["id"] == doc_id:
+            # Simulate loading the full document data
+            st.session_state.current_document = {
+                "id": doc["id"],
+                "filename": doc["filename"],
+                "document_type": doc["document_type"],
+                "original_text": f"Sample content for {doc['filename']}...",  # In real app, load from storage
+                "is_sample": True,  # Mark as sample for demo
+            }
+
+            st.session_state.page = "💬 Q&A"
+            st.rerun()
+            break
+
+
+def delete_document(doc_id: str):
+    """Delete a document from the library."""
+    # Confirm deletion
+    if st.session_state.get(f"confirm_delete_{doc_id}"):
+        documents = st.session_state.get("documents_library", [])
+        st.session_state.documents_library = [
+            doc for doc in documents if doc["id"] != doc_id
+        ]
+
+        # Clear confirmation state
+        del st.session_state[f"confirm_delete_{doc_id}"]
+
+        st.success("✅ Document deleted from library")
+
+
+def show_sample_documents_section():
+    """Show available sample documents for testing."""
+    import os
+
+    st.subheader("🎯 Try Sample Documents")
+    st.markdown("Get started by analyzing our sample legal documents:")
+
+    # Get available sample documents
+    sample_dir = "./sample"
+    sample_files = []
+    if os.path.exists(sample_dir):
+        sample_files = [
+            f for f in os.listdir(sample_dir) if f.endswith((".pdf", ".docx", ".txt"))
+        ]
+
+    if sample_files:
+        # Create description mapping for better UX
+        descriptions = {
+            "Employment_Offer_Letter.pdf": "📋 Analyze employment terms, benefits, and obligations",
+            "Master_Services_Agreement.pdf": "🤝 Review service agreements and contract terms",
+            "Mutual_NDA.pdf": "🔒 Examine confidentiality and non-disclosure clauses",
+            "Residential_Lease_Agreement.pdf": "🏠 Check rental terms, deposits, and tenant rights",
+        }
+
+        for filename in sample_files:
+            with st.expander(
+                f"📄 {filename.replace('_', ' ').replace('.pdf', '')}", expanded=False
+            ):
+                col1, col2 = st.columns([2, 1])
+
+                with col1:
+                    description = descriptions.get(
+                        filename, "📊 Analyze this legal document for risks and terms"
+                    )
+                    st.markdown(description)
+
+                with col2:
+                    if st.button(
+                        "Analyze Now",
+                        key=f"sample_lib_{filename}",
+                        use_container_width=True,
+                    ):
+                        # Set this as the sample to load and redirect to upload page
+                        st.session_state.load_sample = filename
+                        st.session_state.page = "📄 Upload"
+                        st.rerun()
+    else:
+        st.info("No sample documents available.")

src/pages/qa_assistant.py

@@ -0,0 +1,239 @@
+import streamlit as st
+from typing import List, Dict
+import time
+
+from ..services.ai_analyzer import AIAnalyzer
+from ..services.vector_store import VectorStoreService
+
+
+def show_qa_interface():
+    """Display the Q&A assistant interface."""
+
+    if not st.session_state.get("current_document"):
+        st.warning("⚠️ No document loaded. Please upload and analyze a document first.")
+        if st.button("📄 Go to Upload"):
+            st.session_state.page = "📄 Upload"
+            st.rerun()
+        return
+
+    doc = st.session_state.current_document
+
+    # Header
+    st.header("💬 Q&A Assistant")
+    st.markdown(f"Ask questions about **{doc.get('filename', 'your document')}**")
+
+    # Initialize chat history
+    if "qa_history" not in st.session_state:
+        st.session_state.qa_history = []
+
+    # Chat interface
+    chat_container = st.container()
+
+    with chat_container:
+        # Display chat history
+        for i, qa in enumerate(st.session_state.qa_history):
+            # User message
+            with st.chat_message("user"):
+                st.markdown(qa["question"])
+
+            # Assistant response
+            with st.chat_message("assistant"):
+                st.markdown(qa["answer"])
+
+    # Suggested questions
+    st.markdown("### 💡 Suggested Questions")
+
+    doc_type = doc.get("document_type", "other")
+    suggested_questions = get_suggested_questions(doc_type)
+
+    col1, col2 = st.columns(2)
+
+    for i, question in enumerate(suggested_questions):
+        col = col1 if i % 2 == 0 else col2
+        with col:
+            if st.button(question, key=f"suggested_{i}", use_container_width=True):
+                ask_question(question, doc)
+
+    # Chat input
+    st.markdown("### ❓ Ask Your Question")
+
+    with st.form("question_form", clear_on_submit=True):
+        user_question = st.text_input(
+            "Type your question here...",
+            placeholder="e.g., What happens if I terminate this contract early?",
+            label_visibility="collapsed",
+        )
+
+        submitted = st.form_submit_button("Send", use_container_width=True)
+
+        if submitted and user_question.strip():
+            ask_question(user_question, doc)
+
+    # Quick actions
+    st.markdown("---")
+    col1, col2, col3 = st.columns(3)
+
+    with col1:
+        if st.button("📊 Back to Analysis", use_container_width=True):
+            st.session_state.page = "📊 Analysis"
+            st.rerun()
+
+    with col2:
+        if st.button("🗑️ Clear Chat", use_container_width=True):
+            st.session_state.qa_history = []
+            st.rerun()
+
+    with col3:
+        if st.button("📥 Export Chat", use_container_width=True):
+            export_chat_history()
+
+
+def ask_question(question: str, doc: Dict):
+    """Process a question and get AI response."""
+    try:
+        # Show thinking indicator
+        with st.spinner("🤔 Thinking..."):
+            # Initialize AI analyzer
+            ai_analyzer = AIAnalyzer()
+
+            # Get document type
+            from ..models.document import DocumentType
+
+            doc_type = DocumentType(doc.get("document_type", "other"))
+
+            # Get answer from AI
+            answer = ai_analyzer.answer_question(
+                question=question,
+                document_text=doc.get("original_text", ""),
+                document_type=doc_type,
+            )
+
+            # Add to chat history
+            st.session_state.qa_history.append(
+                {"question": question, "answer": answer, "timestamp": time.time()}
+            )
+
+            # Rerun to show the new Q&A
+            st.rerun()
+
+    except Exception as e:
+        st.error(f"❌ Error processing question: {str(e)}")
+
+
+def get_suggested_questions(doc_type: str) -> List[str]:
+    """Get suggested questions based on document type."""
+
+    questions_by_type = {
+        "rental": [
+            "What is the monthly rent amount?",
+            "What happens if I pay rent late?",
+            "How much is the security deposit?",
+            "Can I terminate the lease early?",
+            "Who is responsible for repairs?",
+            "What are the landlord's obligations?",
+            "Are pets allowed in the property?",
+            "What happens if I damage the property?",
+        ],
+        "loan": [
+            "What is the total amount I will repay?",
+            "What is the effective interest rate?",
+            "What happens if I miss a payment?",
+            "What collateral is required?",
+            "Can I repay the loan early?",
+            "What are the processing fees?",
+            "How is the interest calculated?",
+            "What happens in case of default?",
+        ],
+        "employment": [
+            "What is my total compensation package?",
+            "How many hours am I expected to work?",
+            "Can the company terminate me without notice?",
+            "What are the non-compete restrictions?",
+            "Am I allowed to work other jobs?",
+            "What benefits am I entitled to?",
+            "How much notice must I give to resign?",
+            "Who owns the intellectual property I create?",
+        ],
+        "nda": [
+            "What information is considered confidential?",
+            "How long does the confidentiality last?",
+            "What are the penalties for disclosure?",
+            "Can I discuss this agreement with others?",
+            "What happens after the agreement ends?",
+            "Are there any exceptions to confidentiality?",
+        ],
+        "service": [
+            "What services are included in this agreement?",
+            "What is the payment schedule?",
+            "How can this agreement be terminated?",
+            "What are the deliverables and deadlines?",
+            "Who is responsible for what costs?",
+            "What happens if the work is unsatisfactory?",
+        ],
+    }
+
+    return questions_by_type.get(
+        doc_type,
+        [
+            "What are the main obligations for each party?",
+            "What are the key financial terms?",
+            "How can this agreement be terminated?",
+            "What are the potential risks for me?",
+            "What should I be most careful about?",
+            "Are there any unusual or concerning clauses?",
+        ],
+    )
+
+
+def export_chat_history():
+    """Export the chat history as a text file."""
+    if not st.session_state.qa_history:
+        st.warning("No chat history to export.")
+        return
+
+    doc = st.session_state.current_document
+
+    # Create chat export
+    export_text = f"""
+LEGA.AI Q&A SESSION EXPORT
+{'='*50}
+
+Document: {doc.get('filename', 'Unknown')}
+Document Type: {doc.get('document_type', 'Unknown').title()}
+Export Date: {time.strftime('%Y-%m-%d %H:%M:%S')}
+
+QUESTIONS & ANSWERS:
+{'='*50}
+
+"""
+
+    for i, qa in enumerate(st.session_state.qa_history):
+        export_text += f"""
+Q{i+1}: {qa['question']}
+
+A{i+1}: {qa['answer']}
+
+{'-'*30}
+
+"""
+
+    export_text += "\nGenerated by Lega.AI - Making legal documents accessible"
+
+    # Clean filename - remove .pdf extension if present
+    filename = doc.get("filename", "document")
+    if filename.endswith(".pdf"):
+        filename = filename[:-4]
+    if filename.endswith(".docx"):
+        filename = filename[:-5]
+    if filename.endswith(".txt"):
+        filename = filename[:-4]
+
+    # Offer download
+    st.download_button(
+        label="📥 Download Chat History",
+        data=export_text,
+        file_name=f"lega_ai_qa_{filename}.pdf",
+        mime="application/pdf",
+    )
+
+    st.success("✅ Chat history prepared for download!")

src/pages/settings.py

@@ -0,0 +1,304 @@
+import streamlit as st
+from src.utils.config import config
+
+
+def show_settings_interface():
+    """Display the settings interface."""
+
+    st.header("⚙️ Settings")
+    st.markdown("Configure your Lega.AI experience")
+
+    # Tabs for different settings categories
+    tab1, tab2, tab3, tab4 = st.tabs(
+        ["🔑 API Keys", "🎨 Preferences", "📊 Usage", "ℹ️ About"]
+    )
+
+    with tab1:
+        show_api_settings()
+
+    with tab2:
+        show_preference_settings()
+
+    with tab3:
+        show_usage_stats()
+
+    with tab4:
+        show_about_info()
+
+
+def show_api_settings():
+    """Display API key configuration."""
+    st.subheader("🔑 API Configuration")
+
+    # Check current API key status
+    api_key_configured = bool(
+        config.GOOGLE_API_KEY and config.GOOGLE_API_KEY != "your-google-api-key-here"
+    )
+
+    if api_key_configured:
+        st.success("✅ Google AI API key is configured")
+    else:
+        st.warning("⚠️ Google AI API key not configured")
+        st.markdown(
+            """
+        To use Lega.AI's AI features, you need to configure your Google AI API key:
+        
+        1. Go to [Google AI Studio](https://makersuite.google.com/)
+        2. Create a new API key
+        3. Copy the key and add it to your `.env` file
+        4. Set `GOOGLE_API_KEY=your_actual_api_key`
+        5. Restart the application
+        """
+        )
+
+    # API key input (for demonstration)
+    st.markdown("---")
+    st.subheader("🔧 Update API Key")
+
+    with st.form("api_key_form"):
+        new_api_key = st.text_input(
+            "Google AI API Key",
+            type="password",
+            placeholder="Enter your Google AI API key",
+            help="This will be saved to your environment configuration",
+        )
+
+        submitted = st.form_submit_button("Update API Key")
+
+        if submitted:
+            if new_api_key.strip():
+                st.success(
+                    "✅ API key updated! Please restart the application for changes to take effect."
+                )
+                st.info("💡 Don't forget to update your `.env` file with the new key.")
+            else:
+                st.error("❌ Please enter a valid API key")
+
+
+def show_preference_settings():
+    """Display user preference settings."""
+    st.subheader("🎨 User Preferences")
+
+    # Language settings
+    st.markdown("#### 🌐 Language & Region")
+
+    col1, col2 = st.columns(2)
+
+    with col1:
+        language = st.selectbox(
+            "Interface Language",
+            ["English", "Hindi", "Tamil", "Telugu", "Gujarati"],
+            help="Language for the user interface",
+        )
+
+    with col2:
+        region = st.selectbox(
+            "Legal Region",
+            ["India", "Maharashtra", "Delhi", "Karnataka", "Tamil Nadu"],
+            help="Legal jurisdiction for document analysis",
+        )
+
+    # Analysis preferences
+    st.markdown("#### 📊 Analysis Preferences")
+
+    risk_sensitivity = st.slider(
+        "Risk Detection Sensitivity",
+        min_value=1,
+        max_value=5,
+        value=3,
+        help="1 = Only critical risks, 5 = All potential concerns",
+    )
+
+    simplification_level = st.selectbox(
+        "Text Simplification Level",
+        ["Basic", "Intermediate", "Advanced"],
+        index=1,
+        help="How much to simplify legal language",
+    )
+
+    show_suggestions = st.checkbox(
+        "Show improvement suggestions",
+        value=True,
+        help="Display suggestions for problematic clauses",
+    )
+
+    # Notification preferences
+    st.markdown("#### 🔔 Notifications")
+
+    email_notifications = st.checkbox(
+        "Email notifications for analysis completion", value=False
+    )
+
+    browser_notifications = st.checkbox("Browser notifications", value=True)
+
+    # Save preferences
+    if st.button("💾 Save Preferences", type="primary"):
+        # In a real app, save to user profile/database
+        st.success("✅ Preferences saved successfully!")
+
+
+def show_usage_stats():
+    """Display usage statistics."""
+    st.subheader("📊 Usage Statistics")
+
+    # Mock usage data
+    col1, col2, col3 = st.columns(3)
+
+    with col1:
+        st.metric(label="Documents Analyzed", value="47", delta="12 this month")
+
+    with col2:
+        st.metric(label="Questions Asked", value="156", delta="23 this week")
+
+    with col3:
+        st.metric(label="Risks Identified", value="89", delta="High: 12, Medium: 31")
+
+    # Usage by document type
+    st.markdown("#### 📄 Analysis by Document Type")
+
+    usage_data = {
+        "Rental Agreements": 18,
+        "Loan Contracts": 12,
+        "Employment Contracts": 8,
+        "Service Agreements": 6,
+        "NDAs": 3,
+    }
+
+    for doc_type, count in usage_data.items():
+        progress = count / max(usage_data.values())
+        st.markdown(f"**{doc_type}**: {count} documents")
+        st.progress(progress)
+
+    # Storage usage
+    st.markdown("#### 💾 Storage Usage")
+
+    storage_used = 2.4  # GB
+    storage_limit = 5.0  # GB
+
+    st.progress(storage_used / storage_limit)
+    st.markdown(
+        f"**{storage_used:.1f} GB** used of **{storage_limit:.1f} GB** available"
+    )
+
+    # Account tier
+    st.markdown("#### 👤 Account Information")
+
+    col1, col2 = st.columns(2)
+
+    with col1:
+        st.info("**Plan**: Free Tier")
+        st.markdown(
+            """
+        - 10 documents per month
+        - Basic AI analysis
+        - Email support
+        """
+        )
+
+    with col2:
+        st.markdown("**Upgrade Benefits**:")
+        st.markdown(
+            """
+        - Unlimited documents
+        - Advanced AI features
+        - Priority support
+        - Bulk processing
+        """
+        )
+
+        if st.button("🚀 Upgrade to Pro", type="primary"):
+            st.info("Upgrade functionality would be implemented here")
+
+
+def show_about_info():
+    """Display about information."""
+    st.subheader("ℹ️ About Lega.AI")
+
+    # App info
+    col1, col2 = st.columns([2, 1])
+
+    with col1:
+        st.markdown(
+            """
+        **Lega.AI** is an AI-powered platform that makes legal documents accessible to everyone.
+        
+        ### 🎯 Mission
+        To democratize legal document understanding by providing instant AI analysis,
+        risk assessment, and plain language explanations.
+        
+        ### ✨ Features
+        - **Document Analysis**: Upload and analyze any legal document
+        - **Risk Assessment**: Color-coded risk scoring with explanations
+        - **Plain Language**: Convert legal jargon to simple English
+        - **Q&A Assistant**: Ask questions about your documents
+        - **Smart Search**: Find similar clauses and documents
+        - **Export Reports**: Generate comprehensive analysis reports
+        
+        ### 🛡️ Privacy & Security
+        - Your documents are processed securely
+        - No data is shared with third parties
+        - Local vector storage for document similarity
+        - GDPR compliant data handling
+        """
+        )
+
+    with col2:
+        st.markdown(
+            """
+        ### 📊 Version Info
+        **Version**: 1.0.0  
+        **Build**: 2025.09.21  
+        **Engine**: Google Gemini
+        
+        ### 🔧 Tech Stack
+        - **Frontend**: Streamlit
+        - **AI/ML**: LangChain + Gemini
+        - **Vector DB**: Chroma
+        - **Embeddings**: Google Embeddings
+        
+        ### 📞 Support
+        - **Email**: support@lega.ai
+        - **Docs**: github.com/codernoahx/Lega.AI/README.md
+        - **GitHub**: github.com/codernoahx/Lega.AI
+        """
+        )
+
+    # Legal notice
+    st.markdown("---")
+    st.markdown(
+        """
+    ### ⚖️ Legal Notice
+    
+    **Disclaimer**: Lega.AI provides AI-powered analysis for informational purposes only. 
+    This is not legal advice. Always consult with qualified legal professionals for 
+    important legal matters.
+    
+    **Data Usage**: By using this service, you agree to our Terms of Service and Privacy Policy.
+    Your documents are processed to provide analysis but are not used to train AI models.
+    
+    © 2025 Lega.AI. All rights reserved.
+    """
+    )
+
+    # Feedback section
+    st.markdown("---")
+    st.subheader("💬 Feedback")
+
+    with st.form("feedback_form"):
+        feedback_type = st.selectbox(
+            "Feedback Type",
+            ["General Feedback", "Bug Report", "Feature Request", "Question"],
+        )
+
+        feedback_text = st.text_area(
+            "Your Feedback",
+            placeholder="Tell us what you think or report any issues...",
+            height=100,
+        )
+
+        submitted = st.form_submit_button("Send Feedback")
+
+        if submitted and feedback_text.strip():
+            st.success("✅ Thank you for your feedback! We'll review it soon.")
+        elif submitted:
+            st.error("❌ Please enter your feedback before submitting.")

src/pages/upload.py

@@ -0,0 +1,342 @@
+import streamlit as st
+import os
+from typing import Optional
+import time
+
+from ..services.document_processor import DocumentProcessor
+from ..services.ai_analyzer import AIAnalyzer
+from ..services.vector_store import VectorStoreService
+from ..models.document import DocumentType
+from ..utils.helpers import generate_document_id, sanitize_filename, format_file_size
+from ..utils.logger import log_document_upload
+
+
+def show_upload_interface():
+    """Display the document upload interface."""
+    st.header("📄 Upload Legal Document")
+    st.markdown(
+        "Upload your legal document for instant AI analysis and risk assessment."
+    )
+
+    # Check if we should auto-load a sample document
+    if st.session_state.get("load_sample"):
+        filename = st.session_state.load_sample
+        del st.session_state.load_sample  # Clear the flag
+        load_sample_document_from_file(filename)
+        return
+
+    # File uploader
+    uploaded_file = st.file_uploader(
+        "Choose a file",
+        type=["pdf", "txt", "docx"],  # Added docx support
+        help="Supported formats: PDF, TXT, DOCX (Max 10MB)",
+        key="document_uploader",
+    )
+
+    if uploaded_file is not None:
+        # Display file info
+        file_size = len(uploaded_file.getvalue())
+
+        # Check file size limit
+        max_size = 10 * 1024 * 1024  # 10MB
+        if file_size > max_size:
+            st.error(f"❌ File too large. Maximum size is {format_file_size(max_size)}")
+            return
+
+        st.success(f"📁 **{uploaded_file.name}** ({format_file_size(file_size)})")
+
+        # Process button
+        if st.button("🔍 Analyze Document", type="primary", use_container_width=True):
+            process_uploaded_document(uploaded_file)
+
+    # Sample documents section
+    st.markdown("---")
+    st.subheader("📋 Try Sample Documents")
+    st.markdown("Don't have a document handy? Try one of our real sample documents:")
+
+    # Get available sample documents
+    sample_dir = "./sample"
+    sample_files = []
+    if os.path.exists(sample_dir):
+        sample_files = [f for f in os.listdir(sample_dir) if f.endswith(('.pdf', '.docx', '.txt'))]
+
+    if sample_files:
+        col1, col2 = st.columns(2)
+        
+        for i, filename in enumerate(sample_files):
+            col = col1 if i % 2 == 0 else col2
+            
+            with col:
+                # Create descriptive button names
+                display_name = filename.replace('_', ' ').replace('.pdf', '').replace('.docx', '').replace('.txt', '')
+                display_name = display_name.title()
+                
+                if st.button(f"📄 {display_name}", use_container_width=True, key=f"sample_{i}"):
+                    load_sample_document_from_file(filename)
+    else:
+        st.info("No sample documents found in the sample directory.")
+
+
+def process_uploaded_document(uploaded_file):
+    """Process the uploaded document with AI analysis."""
+    try:
+        # Initialize processors
+        doc_processor = DocumentProcessor()
+        ai_analyzer = AIAnalyzer()
+        vector_store = VectorStoreService()
+
+        # Create progress tracking
+        progress_bar = st.progress(0)
+        status_text = st.empty()
+
+        # Step 1: Extract text
+        status_text.text("📄 Extracting text from document...")
+        progress_bar.progress(20)
+
+        file_content = uploaded_file.getvalue()
+        text = doc_processor.extract_text(file_content, uploaded_file.name)
+
+        if not text.strip():
+            st.error(
+                "❌ Could not extract text from the document. Please try a different file."
+            )
+            progress_bar.empty()
+            status_text.empty()
+            return
+
+        progress_bar.progress(40)
+
+        # Step 2: Detect document type
+        status_text.text("🔍 Analyzing document type...")
+        document_type = doc_processor.detect_document_type(text)
+        progress_bar.progress(50)
+
+        # Step 3: Risk analysis
+        status_text.text("⚠️ Performing risk assessment...")
+        risk_data = ai_analyzer.analyze_document_risk(text, document_type)
+        progress_bar.progress(70)
+
+        # Step 4: Text simplification
+        status_text.text("💬 Simplifying legal language...")
+        simplified_data = ai_analyzer.simplify_text(text, document_type)
+        progress_bar.progress(85)
+
+        # Step 5: Generate summary
+        status_text.text("📋 Generating summary...")
+        summary = ai_analyzer.generate_summary(text, document_type)
+
+        # Step 6: Add to vector store
+        status_text.text("💾 Storing document for search...")
+        doc_id = generate_document_id()
+        vector_store.add_document(
+            document_id=doc_id,
+            text=text,
+            metadata={
+                "filename": uploaded_file.name,
+                "document_type": document_type.value,
+                "upload_date": time.strftime("%Y-%m-%d %H:%M:%S"),
+            },
+        )
+
+        progress_bar.progress(100)
+
+        # Complete
+        status_text.text("✅ Analysis complete!")
+        time.sleep(1)
+        progress_bar.empty()
+        status_text.empty()
+
+        # Store results in session state
+        st.session_state.current_document = {
+            "id": doc_id,
+            "filename": uploaded_file.name,
+            "document_type": document_type.value,
+            "original_text": text,
+            "simplified_text": simplified_data.get("simplified_text", ""),
+            "summary": summary,
+            "risk_data": risk_data,
+            "key_points": simplified_data.get("key_points", []),
+            "jargon_definitions": simplified_data.get("jargon_definitions", {}),
+            "analysis_timestamp": time.time(),
+            "file_size": len(file_content),
+        }
+
+        # Add to documents library
+        if "documents_library" not in st.session_state:
+            st.session_state.documents_library = []
+
+        st.session_state.documents_library.append(
+            {
+                "id": doc_id,
+                "filename": uploaded_file.name,
+                "document_type": document_type.value,
+                "upload_date": time.strftime("%Y-%m-%d %H:%M:%S"),
+                "file_size": len(file_content),
+                "risk_score": len(risk_data.get("risk_factors", []))
+                * 10,  # Simple risk score
+            }
+        )
+
+        # Log the upload
+        log_document_upload(uploaded_file.name, len(file_content))
+
+        # Show success and redirect to analysis page
+        st.success("🎉 Document analysis completed! Redirecting to results...")
+
+        # Set page state for redirection
+        st.session_state.page = "📊 Analysis"
+
+        time.sleep(2)
+        st.rerun()
+
+    except Exception as e:
+        st.error(f"❌ Error processing document: {str(e)}")
+        progress_bar.empty()
+        status_text.empty()
+
+
+def load_sample_document_from_file(filename: str):
+    """Load an actual sample document from the sample directory."""
+    try:
+        sample_path = os.path.join("./sample", filename)
+        
+        if not os.path.exists(sample_path):
+            st.error(f"❌ Sample file not found: {filename}")
+            return
+        
+        # Read the file
+        with open(sample_path, 'rb') as f:
+            file_content = f.read()
+        
+        # Create a mock uploaded file object
+        class MockUploadedFile:
+            def __init__(self, content, name):
+                self._content = content
+                self.name = name
+            
+            def getvalue(self):
+                return self._content
+        
+        mock_file = MockUploadedFile(file_content, filename)
+        
+        st.success(f"📄 Loading sample document: **{filename}**")
+        
+        # Process the sample document
+        process_uploaded_document(mock_file)
+        
+    except Exception as e:
+        st.error(f"❌ Error loading sample document: {str(e)}")
+
+
+def load_sample_document(doc_type: str):
+    """Load a sample document for demonstration."""
+    sample_docs = {
+        "rental": {
+            "filename": "sample_rental_agreement.pdf",
+            "type": "rental",
+            "text": """
+            RESIDENTIAL LEASE AGREEMENT
+            
+            This Lease Agreement is entered into between John Smith (Landlord) and Jane Doe (Tenant) 
+            for the property located at 123 Main Street, Mumbai, Maharashtra.
+            
+            RENT: Tenant agrees to pay Rs. 25,000 per month, due on the 1st of each month. 
+            Late payments will incur a penalty of Rs. 1,000 per day.
+            
+            SECURITY DEPOSIT: Tenant shall pay a security deposit of Rs. 75,000, which is 
+            non-refundable except for damage assessment.
+            
+            TERMINATION: Either party may terminate this lease with 30 days written notice. 
+            Early termination by Tenant results in forfeiture of security deposit.
+            
+            MAINTENANCE: Tenant is responsible for all repairs and maintenance, including 
+            structural repairs, regardless of cause.
+            
+            The property is leased "as-is" with no warranties. Landlord is not liable for 
+            any damages or injuries occurring on the premises.
+            """,
+        },
+        "loan": {
+            "filename": "sample_loan_agreement.pdf",
+            "type": "loan",
+            "text": """
+            PERSONAL LOAN AGREEMENT
+            
+            Borrower: Rajesh Kumar
+            Lender: QuickCash Financial Services Pvt Ltd
+            Principal Amount: Rs. 2,00,000
+            
+            INTEREST RATE: 24% per annum (APR 28.5% including processing fees)
+            
+            REPAYMENT: 24 monthly installments of Rs. 12,500 each
+            Total repayment amount: Rs. 3,00,000
+            
+            LATE PAYMENT PENALTY: Rs. 500 per day for any late payment
+            
+            DEFAULT: If payment is late by more than 7 days, the entire remaining 
+            balance becomes immediately due and payable.
+            
+            COLLATERAL: Borrower pledges gold ornaments worth Rs. 2,50,000 as security.
+            Lender may seize collateral immediately upon default.
+            
+            ARBITRATION: All disputes shall be resolved through binding arbitration. 
+            Borrower waives right to jury trial.
+            
+            Processing fee: Rs. 10,000 (non-refundable)
+            Documentation charges: Rs. 5,000
+            """,
+        },
+        "employment": {
+            "filename": "sample_employment_contract.docx",  # Changed to DOCX
+            "type": "employment",
+            "text": """
+            EMPLOYMENT CONTRACT
+            
+            Employee: Priya Sharma
+            Company: TechCorp India Private Limited
+            Position: Software Developer
+            Start Date: January 1, 2024
+            
+            SALARY: Rs. 8,00,000 per annum, payable monthly
+            
+            WORKING HOURS: 45 hours per week, including mandatory weekend work when required
+            
+            NON-COMPETE: Employee shall not work for any competing company for 2 years 
+            after termination, within India or globally.
+            
+            CONFIDENTIALITY: Employee agrees to maintain strict confidentiality of all 
+            company information indefinitely, even after termination.
+            
+            TERMINATION: Company may terminate employment at any time without cause or notice. 
+            Employee must provide 90 days notice to resign.
+            
+            NO MOONLIGHTING: Employee shall not engage in any other work or business 
+            activities during employment.
+            
+            INTELLECTUAL PROPERTY: All work created by employee belongs entirely to company, 
+            including personal projects done outside work hours.
+            """,
+        },
+    }
+
+    if doc_type in sample_docs:
+        sample = sample_docs[doc_type]
+        from ..utils.helpers import generate_document_id
+
+        # Store in session state
+        doc_id = generate_document_id()
+        st.session_state.current_document = {
+            "id": doc_id,
+            "filename": sample["filename"],
+            "document_type": sample["type"],
+            "original_text": sample["text"],
+            "is_sample": True,
+        }
+
+        st.success(f"📄 Loaded sample {doc_type} document. Processing...")
+
+        # Simulate processing for demo
+        with st.spinner("Analyzing sample document..."):
+            time.sleep(2)
+
+        st.rerun()

src/services/__init__.py

@@ -0,0 +1 @@
+# src/services/__init__.py

src/services/ai_analyzer.py

@@ -0,0 +1,378 @@
+from typing import List, Dict, Any, Optional
+from langchain_google_genai import ChatGoogleGenerativeAI
+from langchain.prompts import PromptTemplate
+from langchain.chains import LLMChain
+import json
+import time
+
+from src.utils.config import config
+from src.utils.logger import log_error, log_analysis_start, log_analysis_complete
+from src.models.document import (
+    DocumentType,
+    RiskLevel,
+    RiskCategory,
+    RiskFactor,
+    ClausePosition,
+)
+from src.utils.helpers import (
+    calculate_risk_score,
+    extract_key_dates,
+    extract_financial_terms,
+)
+
+
+class AIAnalyzer:
+    def __init__(self):
+        self.llm = ChatGoogleGenerativeAI(
+            model=config.CHAT_MODEL,
+            google_api_key=config.GOOGLE_API_KEY,
+            temperature=config.TEMPERATURE,
+            max_output_tokens=config.MAX_TOKENS,
+        )
+
+        # Initialize prompt templates
+        self._setup_prompts()
+
+    def _setup_prompts(self):
+        """Set up prompt templates for different analysis tasks."""
+
+        # Risk analysis prompt
+        self.risk_analysis_prompt = PromptTemplate(
+            input_variables=["text", "document_type"],
+            template="""
+            Analyze the following {document_type} document for potential risks and problematic clauses.
+            
+            Document text:
+            {text}
+            
+            Please identify:
+            1. High-risk clauses that could be problematic for the signer
+            2. Financial risks (hidden fees, penalties, high costs)
+            3. Commitment risks (long-term obligations, difficult exit clauses)
+            4. Rights risks (waived protections, limited recourse)
+            
+            For each risk, provide:
+            - The exact clause text (keep it concise, max 100 words)
+            - Risk category (financial, commitment, rights, or standard)
+            - Severity level (low, medium, high, critical)
+            - Clear explanation of why it's risky
+            - Suggestion for improvement
+            
+            IMPORTANT: Return ONLY valid JSON in the exact format below. Do not include any explanatory text before or after the JSON:
+            
+            {{
+                "risk_factors": [
+                    {{
+                        "clause_text": "exact text from document",
+                        "category": "financial",
+                        "severity": "medium",
+                        "explanation": "why this is risky",
+                        "suggestion": "how to improve or what to watch for"
+                    }}
+                ],
+                "overall_assessment": "brief summary of document risk level"
+            }}
+            """,
+        )
+
+        # Plain language translation prompt
+        self.simplification_prompt = PromptTemplate(
+            input_variables=["text", "document_type"],
+            template="""
+            Convert the following legal text into plain, simple English that anyone can understand.
+            
+            Document type: {document_type}
+            Legal text: {text}
+            
+            Rules for simplification:
+            1. Use everyday language instead of legal jargon
+            2. Break down complex sentences into shorter ones
+            3. Explain what actions or obligations mean in practical terms
+            4. Keep the essential meaning intact
+            5. Use "you" to make it personal and clear
+            6. Focus on the most important points
+            
+            IMPORTANT: Return ONLY valid JSON in the exact format below. Do not include any explanatory text:
+            
+            {{
+                "simplified_text": "the simplified version in plain English",
+                "key_points": ["main point 1", "main point 2", "main point 3"],
+                "jargon_definitions": {{"legal term": "simple definition"}}
+            }}
+            """,
+        )
+
+        # Document summary prompt
+        self.summary_prompt = PromptTemplate(
+            input_variables=["text", "document_type"],
+            template="""
+            Create a concise summary of this {document_type} document.
+            
+            Document: {text}
+            
+            Provide a summary that includes:
+            1. What type of agreement this is
+            2. Who are the main parties involved
+            3. Key obligations for each party
+            4. Important terms (dates, amounts, conditions)
+            5. Major benefits and risks
+            
+            Keep it under 200 words and focus on what matters most to the person signing.
+            """,
+        )
+
+    def analyze_document_risk(
+        self, text: str, document_type: DocumentType
+    ) -> Dict[str, Any]:
+        """Analyze document for risks and problematic clauses."""
+        try:
+            log_analysis_start("risk_analysis")
+            start_time = time.time()
+
+            # Create and run the risk analysis chain
+            risk_chain = LLMChain(llm=self.llm, prompt=self.risk_analysis_prompt)
+
+            result = risk_chain.run(
+                text=text[:4000],  # Limit text size for API
+                document_type=document_type.value,
+            )
+
+            # Parse JSON response with better error handling
+            try:
+                # Try to extract JSON from the response if it's wrapped in markdown
+                if "```json" in result:
+                    json_start = result.find("```json") + 7
+                    json_end = result.find("```", json_start)
+                    if json_end != -1:
+                        result = result[json_start:json_end].strip()
+
+                # Clean up the result string
+                result = result.strip()
+                if result.startswith("```") and result.endswith("```"):
+                    result = result[3:-3].strip()
+
+                risk_data = json.loads(result)
+
+                # Validate the structure
+                if not isinstance(risk_data, dict):
+                    raise ValueError("Response is not a dictionary")
+
+                if "risk_factors" not in risk_data:
+                    risk_data["risk_factors"] = []
+
+                if "overall_assessment" not in risk_data:
+                    risk_data["overall_assessment"] = "Analysis completed"
+
+            except (json.JSONDecodeError, ValueError) as e:
+                log_error(f"JSON parsing error in risk analysis: {str(e)}")
+                log_error(f"Raw AI response: {result[:500]}...")
+
+                # Try to extract risk information manually if JSON fails
+                risk_data = self._extract_risk_fallback(result, text)
+
+            processing_time = time.time() - start_time
+            log_analysis_complete("risk_analysis", processing_time)
+
+            return risk_data
+
+        except Exception as e:
+            log_error(f"Error in risk analysis: {str(e)}")
+            return {"risk_factors": [], "overall_assessment": "Analysis failed"}
+
+    def _extract_risk_fallback(
+        self, response: str, original_text: str
+    ) -> Dict[str, Any]:
+        """Fallback method to extract risk information when JSON parsing fails."""
+        try:
+            risk_factors = []
+
+            # Look for common risk indicators in the response
+            risk_keywords = [
+                "risk",
+                "problematic",
+                "concern",
+                "warning",
+                "caution",
+                "penalty",
+                "fee",
+            ]
+            sentences = response.split(".")
+
+            for i, sentence in enumerate(sentences):
+                sentence = sentence.strip()
+                if (
+                    any(
+                        keyword.lower() in sentence.lower() for keyword in risk_keywords
+                    )
+                    and len(sentence) > 20
+                ):
+                    risk_factors.append(
+                        {
+                            "clause_text": sentence[:200],  # Limit length
+                            "category": "standard",
+                            "severity": "medium",
+                            "explanation": "Potential risk identified by text analysis",
+                            "suggestion": "Review this clause carefully with legal counsel",
+                        }
+                    )
+
+                    if len(risk_factors) >= 5:  # Limit to 5 fallback risks
+                        break
+
+            return {
+                "risk_factors": risk_factors,
+                "overall_assessment": "Risk analysis completed with limited parsing. Please review manually.",
+            }
+
+        except Exception as e:
+            log_error(f"Error in fallback risk extraction: {str(e)}")
+            return {
+                "risk_factors": [],
+                "overall_assessment": "Unable to analyze risks - please try again",
+            }
+
+    def simplify_text(self, text: str, document_type: DocumentType) -> Dict[str, Any]:
+        """Convert legal text to plain language."""
+        try:
+            simplification_chain = LLMChain(
+                llm=self.llm, prompt=self.simplification_prompt
+            )
+
+            result = simplification_chain.run(
+                text=text[:3000], document_type=document_type.value  # Limit text size
+            )
+
+            # Parse JSON response with better error handling
+            try:
+                # Try to extract JSON from the response if it's wrapped in markdown
+                if "```json" in result:
+                    json_start = result.find("```json") + 7
+                    json_end = result.find("```", json_start)
+                    if json_end != -1:
+                        result = result[json_start:json_end].strip()
+
+                # Clean up the result string
+                result = result.strip()
+                if result.startswith("```") and result.endswith("```"):
+                    result = result[3:-3].strip()
+
+                simplified_data = json.loads(result)
+
+                # Validate the structure
+                if not isinstance(simplified_data, dict):
+                    raise ValueError("Response is not a dictionary")
+
+                # Ensure required keys exist
+                if "simplified_text" not in simplified_data:
+                    simplified_data["simplified_text"] = text[:500] + "..."
+                if "key_points" not in simplified_data:
+                    simplified_data["key_points"] = ["Unable to extract key points"]
+                if "jargon_definitions" not in simplified_data:
+                    simplified_data["jargon_definitions"] = {}
+
+            except (json.JSONDecodeError, ValueError) as e:
+                log_error(f"JSON parsing error in text simplification: {str(e)}")
+                simplified_data = {
+                    "simplified_text": text[:500]
+                    + "... (Full simplification unavailable)",
+                    "key_points": ["Document content requires legal review"],
+                    "jargon_definitions": {},
+                }
+
+            return simplified_data
+
+        except Exception as e:
+            log_error(f"Error in text simplification: {str(e)}")
+            return {
+                "simplified_text": text[:500]
+                + "...",  # Return truncated original if simplification fails
+                "key_points": ["Simplification failed - showing original text"],
+                "jargon_definitions": {},
+            }
+
+    def generate_summary(self, text: str, document_type: DocumentType) -> str:
+        """Generate a concise document summary."""
+        try:
+            summary_chain = LLMChain(llm=self.llm, prompt=self.summary_prompt)
+
+            summary = summary_chain.run(
+                text=text[:3000], document_type=document_type.value  # Limit text size
+            )
+
+            return summary.strip()
+
+        except Exception as e:
+            log_error(f"Error generating summary: {str(e)}")
+            return "Unable to generate summary"
+
+    def answer_question(
+        self, question: str, document_text: str, document_type: DocumentType
+    ) -> str:
+        """Answer a question about the document."""
+        try:
+            qa_prompt = PromptTemplate(
+                input_variables=["question", "document", "doc_type"],
+                template="""
+                Answer the following question about this {doc_type} document. 
+                Be specific and cite relevant parts of the document.
+                
+                Document: {document}
+                
+                Question: {question}
+                
+                Provide a clear, helpful answer based only on the document content.
+                If the answer isn't in the document, say so clearly.
+                """,
+            )
+
+            qa_chain = LLMChain(llm=self.llm, prompt=qa_prompt)
+
+            answer = qa_chain.run(
+                question=question,
+                document=document_text[:3000],  # Limit context size
+                doc_type=document_type.value,
+            )
+
+            return answer.strip()
+
+        except Exception as e:
+            log_error(f"Error answering question: {str(e)}")
+            return "Sorry, I couldn't process your question. Please try again."
+
+    def create_risk_factors(
+        self, risk_data: Dict[str, Any], text: str
+    ) -> List[RiskFactor]:
+        """Convert AI analysis results to RiskFactor objects."""
+        risk_factors = []
+
+        for factor_data in risk_data.get("risk_factors", []):
+            try:
+                # Find clause position in text
+                clause_text = factor_data.get("clause_text", "")
+                position = None
+
+                if clause_text:
+                    start_index = text.find(clause_text)
+                    if start_index != -1:
+                        position = ClausePosition(
+                            start_index=start_index,
+                            end_index=start_index + len(clause_text),
+                        )
+
+                risk_factor = RiskFactor(
+                    id=f"risk_{len(risk_factors) + 1}",
+                    clause_text=clause_text,
+                    category=RiskCategory(factor_data.get("category", "standard")),
+                    severity=RiskLevel(factor_data.get("severity", "low")),
+                    explanation=factor_data.get("explanation", ""),
+                    suggestion=factor_data.get("suggestion"),
+                    position=position,
+                )
+
+                risk_factors.append(risk_factor)
+
+            except Exception as e:
+                log_error(f"Error creating risk factor: {str(e)}")
+                continue
+
+        return risk_factors

src/services/document_processor.py

@@ -0,0 +1,202 @@
+from typing import BinaryIO, Optional
+from langchain_community.document_loaders import PyPDFLoader, TextLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+import tempfile
+import os
+from docx import Document
+
+from src.utils.config import config
+from src.utils.logger import log_error
+from src.models.document import DocumentType
+
+
+class DocumentProcessor:
+    def __init__(self):
+        self.text_splitter = RecursiveCharacterTextSplitter(
+            chunk_size=1000,
+            chunk_overlap=200,
+            length_function=len,
+        )
+
+    def extract_text_from_pdf(self, file_content: bytes, filename: str) -> str:
+        """Extract text from PDF using LangChain PyPDFLoader."""
+        try:
+            # Save uploaded file to temporary location
+            with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as temp_file:
+                temp_file.write(file_content)
+                temp_file_path = temp_file.name
+
+            # Use LangChain PyPDFLoader
+            loader = PyPDFLoader(temp_file_path)
+            documents = loader.load()
+
+            # Combine all pages
+            text = "\n".join([doc.page_content for doc in documents])
+
+            # Clean up temporary file
+            os.unlink(temp_file_path)
+
+            return text
+
+        except Exception as e:
+            log_error(f"Error extracting text from PDF: {str(e)}")
+            return ""
+
+    def extract_text_from_txt(self, file_content: bytes, filename: str) -> str:
+        """Extract text from TXT file."""
+        try:
+            # Try different encodings
+            encodings = ["utf-8", "utf-16", "latin-1", "cp1252"]
+
+            for encoding in encodings:
+                try:
+                    text = file_content.decode(encoding)
+                    return text
+                except UnicodeDecodeError:
+                    continue
+
+            # If all encodings fail, use utf-8 with error handling
+            return file_content.decode("utf-8", errors="ignore")
+
+        except Exception as e:
+            log_error(f"Error extracting text from TXT: {str(e)}")
+            return ""
+
+    def extract_text_from_docx(self, file_content: bytes, filename: str) -> str:
+        """Extract text from DOCX file."""
+        try:
+            # Save uploaded file to temporary location
+            with tempfile.NamedTemporaryFile(delete=False, suffix=".docx") as temp_file:
+                temp_file.write(file_content)
+                temp_file_path = temp_file.name
+
+            # Use python-docx to extract text
+            from docx import Document as DocxDocument
+
+            doc = DocxDocument(temp_file_path)
+
+            # Extract text from all paragraphs
+            text_parts = []
+            for paragraph in doc.paragraphs:
+                if paragraph.text.strip():
+                    text_parts.append(paragraph.text)
+
+            # Extract text from tables
+            for table in doc.tables:
+                for row in table.rows:
+                    for cell in row.cells:
+                        if cell.text.strip():
+                            text_parts.append(cell.text)
+
+            # Clean up temporary file
+            os.unlink(temp_file_path)
+
+            # Join all text parts
+            full_text = "\n".join(text_parts)
+            return full_text
+
+        except Exception as e:
+            log_error(f"Error extracting text from DOCX: {str(e)}")
+            return ""
+
+    def extract_text(self, file_content: bytes, filename: str) -> str:
+        """Extract text based on file extension."""
+        file_ext = filename.lower().split(".")[-1]
+
+        if file_ext == "pdf":
+            return self.extract_text_from_pdf(file_content, filename)
+        elif file_ext == "txt":
+            return self.extract_text_from_txt(file_content, filename)
+        elif file_ext in ["docx", "doc"]:
+            return self.extract_text_from_docx(file_content, filename)
+        else:
+            log_error(f"Unsupported file type: {file_ext}")
+            return ""
+
+    def split_text_into_chunks(self, text: str) -> list:
+        """Split text into manageable chunks for processing."""
+        return self.text_splitter.split_text(text)
+
+    def detect_document_type(self, text: str) -> DocumentType:
+        """Detect document type based on content."""
+        text_lower = text.lower()
+
+        # Rental agreement keywords
+        rental_keywords = [
+            "lease",
+            "rent",
+            "tenant",
+            "landlord",
+            "property",
+            "premises",
+            "deposit",
+        ]
+
+        # Loan agreement keywords
+        loan_keywords = [
+            "loan",
+            "borrow",
+            "lender",
+            "principal",
+            "interest",
+            "repayment",
+            "credit",
+        ]
+
+        # Employment keywords
+        employment_keywords = [
+            "employment",
+            "employee",
+            "employer",
+            "salary",
+            "wages",
+            "position",
+            "job",
+        ]
+
+        # NDA keywords
+        nda_keywords = ["confidential", "non-disclosure", "proprietary", "trade secret"]
+
+        # Service agreement keywords
+        service_keywords = [
+            "service",
+            "provider",
+            "client",
+            "deliverables",
+            "scope of work",
+        ]
+
+        # Count keyword matches
+        scores = {
+            DocumentType.RENTAL: sum(
+                1 for keyword in rental_keywords if keyword in text_lower
+            ),
+            DocumentType.LOAN: sum(
+                1 for keyword in loan_keywords if keyword in text_lower
+            ),
+            DocumentType.EMPLOYMENT: sum(
+                1 for keyword in employment_keywords if keyword in text_lower
+            ),
+            DocumentType.NDA: sum(
+                1 for keyword in nda_keywords if keyword in text_lower
+            ),
+            DocumentType.SERVICE: sum(
+                1 for keyword in service_keywords if keyword in text_lower
+            ),
+        }
+
+        # Return type with highest score, or OTHER if no clear match
+        if max(scores.values()) > 2:
+            return max(scores, key=scores.get)
+        else:
+            return DocumentType.OTHER
+
+    def extract_metadata(self, text: str) -> dict:
+        """Extract metadata from document text."""
+        metadata = {
+            "word_count": len(text.split()),
+            "character_count": len(text),
+            "estimated_reading_time": len(text.split()) // 200,  # Assuming 200 WPM
+        }
+
+        return metadata

src/services/vector_store.py

@@ -0,0 +1,212 @@
+from typing import List, Dict, Any, Optional
+from langchain_google_genai import GoogleGenerativeAIEmbeddings
+from langchain_chroma import Chroma
+from langchain.schema import Document
+import os
+
+from src.utils.config import config
+from src.utils.logger import log_error
+from src.models.document import Document as DocModel
+
+
+class VectorStoreService:
+    def __init__(self):
+        # Initialize embeddings
+        self.embeddings = GoogleGenerativeAIEmbeddings(
+            model=config.EMBEDDING_MODEL, google_api_key=config.GOOGLE_API_KEY
+        )
+
+        # Ensure Chroma directory exists
+        os.makedirs(config.CHROMA_PERSIST_DIR, exist_ok=True)
+
+        # Initialize Chroma vector store
+        self.vector_store = Chroma(
+            persist_directory=config.CHROMA_PERSIST_DIR,
+            embedding_function=self.embeddings,
+            collection_name="lega_documents",
+        )
+
+    def add_document(
+        self, document_id: str, text: str, metadata: Dict[str, Any] = None
+    ) -> bool:
+        """Add a document to the vector store."""
+        try:
+            # Create document chunks for better retrieval
+            chunks = self._chunk_document(text)
+
+            documents = []
+            metadatas = []
+            ids = []
+
+            for i, chunk in enumerate(chunks):
+                chunk_metadata = {
+                    "document_id": document_id,
+                    "chunk_id": i,
+                    "chunk_type": "text",
+                    **(metadata or {}),
+                }
+
+                documents.append(chunk)
+                metadatas.append(chunk_metadata)
+                ids.append(f"{document_id}_chunk_{i}")
+
+            # Add to vector store
+            self.vector_store.add_texts(texts=documents, metadatas=metadatas, ids=ids)
+
+            return True
+
+        except Exception as e:
+            log_error(f"Error adding document to vector store: {str(e)}")
+            return False
+
+    def search_similar_documents(self, query: str, k: int = 5) -> List[Dict[str, Any]]:
+        """Search for similar documents based on query."""
+        try:
+            results = self.vector_store.similarity_search_with_score(query=query, k=k)
+
+            formatted_results = []
+            for doc, score in results:
+                formatted_results.append(
+                    {
+                        "content": doc.page_content,
+                        "metadata": doc.metadata,
+                        "similarity_score": score,
+                    }
+                )
+
+            return formatted_results
+
+        except Exception as e:
+            log_error(f"Error searching vector store: {str(e)}")
+            return []
+
+    def search_document_clauses(
+        self, document_id: str, query: str, k: int = 3
+    ) -> List[Dict[str, Any]]:
+        """Search for specific clauses within a document."""
+        try:
+            # Filter by document_id
+            results = self.vector_store.similarity_search_with_score(
+                query=query, k=k, filter={"document_id": document_id}
+            )
+
+            formatted_results = []
+            for doc, score in results:
+                formatted_results.append(
+                    {
+                        "content": doc.page_content,
+                        "metadata": doc.metadata,
+                        "similarity_score": score,
+                    }
+                )
+
+            return formatted_results
+
+        except Exception as e:
+            log_error(f"Error searching document clauses: {str(e)}")
+            return []
+
+    def get_document_context(
+        self, document_id: str, query: str, max_chunks: int = 5
+    ) -> str:
+        """Get relevant context from a document for Q&A."""
+        try:
+            results = self.search_document_clauses(document_id, query, k=max_chunks)
+
+            # Combine relevant chunks
+            context_parts = []
+            for result in results:
+                if result["similarity_score"] < 0.8:  # Only use highly relevant chunks
+                    context_parts.append(result["content"])
+
+            return "\n\n".join(context_parts)
+
+        except Exception as e:
+            log_error(f"Error getting document context: {str(e)}")
+            return ""
+
+    def remove_document(self, document_id: str) -> bool:
+        """Remove a document and all its chunks from the vector store."""
+        try:
+            # Get all chunks for this document
+            results = self.vector_store.get(where={"document_id": document_id})
+
+            if results and results.get("ids"):
+                # Delete all chunks
+                self.vector_store.delete(ids=results["ids"])
+
+            return True
+
+        except Exception as e:
+            log_error(f"Error removing document from vector store: {str(e)}")
+            return False
+
+    def get_document_stats(self) -> Dict[str, Any]:
+        """Get statistics about the vector store."""
+        try:
+            # Get collection info
+            collection = self.vector_store._collection
+            count = collection.count()
+
+            return {
+                "total_documents": count,
+                "collection_name": "lega_documents",
+                "persist_directory": config.CHROMA_PERSIST_DIR,
+            }
+
+        except Exception as e:
+            log_error(f"Error getting vector store stats: {str(e)}")
+            return {"total_documents": 0}
+
+    def _chunk_document(
+        self, text: str, chunk_size: int = 1000, overlap: int = 200
+    ) -> List[str]:
+        """Split document into chunks for embedding."""
+        chunks = []
+        start = 0
+
+        while start < len(text):
+            end = start + chunk_size
+            chunk = text[start:end]
+
+            # Try to break at sentence boundary
+            if end < len(text):
+                last_period = chunk.rfind(".")
+                if last_period > chunk_size // 2:
+                    chunk = chunk[: last_period + 1]
+                    end = start + last_period + 1
+
+            chunks.append(chunk.strip())
+            start = end - overlap
+
+        return [chunk for chunk in chunks if chunk.strip()]
+
+    def find_similar_clauses(
+        self, clause_text: str, exclude_document_id: str = None, k: int = 3
+    ) -> List[Dict[str, Any]]:
+        """Find similar clauses across all documents."""
+        try:
+            filter_dict = {}
+            if exclude_document_id:
+                # This is a simplified filter - Chroma might need different syntax
+                filter_dict = {"document_id": {"$ne": exclude_document_id}}
+
+            results = self.vector_store.similarity_search_with_score(
+                query=clause_text, k=k, filter=filter_dict if filter_dict else None
+            )
+
+            formatted_results = []
+            for doc, score in results:
+                formatted_results.append(
+                    {
+                        "content": doc.page_content,
+                        "metadata": doc.metadata,
+                        "similarity_score": score,
+                    }
+                )
+
+            return formatted_results
+
+        except Exception as e:
+            log_error(f"Error finding similar clauses: {str(e)}")
+            return []

src/utils/__init__.py

@@ -0,0 +1 @@
+# src/utils/__init__.py

src/utils/config.py

@@ -0,0 +1,175 @@
+import os
+from typing import Optional
+from dotenv import load_dotenv
+
+# Load environment variables
+load_dotenv()
+
+
+class Config:
+    # =============================================================================
+    # GOOGLE AI API CONFIGURATION
+    # =============================================================================
+    GOOGLE_API_KEY: str = os.getenv("GOOGLE_API_KEY", "")
+
+    # =============================================================================
+    # APPLICATION SETTINGS
+    # =============================================================================
+    DEBUG: bool = os.getenv("DEBUG", "True").lower() == "true"
+    STREAMLIT_SERVER_PORT: int = int(os.getenv("STREAMLIT_SERVER_PORT", "8501"))
+    STREAMLIT_SERVER_ADDRESS: str = os.getenv("STREAMLIT_SERVER_ADDRESS", "localhost")
+    MAX_FILE_SIZE_MB: int = int(os.getenv("MAX_FILE_SIZE_MB", "10"))
+    SUPPORTED_FILE_TYPES: list = os.getenv(
+        "SUPPORTED_FILE_TYPES", "pdf,docx,txt"
+    ).split(",")
+
+    # =============================================================================
+    # LOGGING
+    # =============================================================================
+    LOG_LEVEL: str = os.getenv("LOG_LEVEL", "INFO")
+    LOG_FILE: str = os.getenv("LOG_FILE", "./data/app.log")
+
+    # =============================================================================
+    # SECURITY
+    # =============================================================================
+    SECRET_KEY: str = os.getenv("SECRET_KEY", "development-key-change-in-production")
+    SESSION_TIMEOUT_MINUTES: int = int(os.getenv("SESSION_TIMEOUT_MINUTES", "60"))
+
+    # =============================================================================
+    # AI MODEL SETTINGS
+    # =============================================================================
+    CHAT_MODEL: str = os.getenv("CHAT_MODEL", "gemini-1.5-flash")
+    TEMPERATURE: float = float(os.getenv("TEMPERATURE", "0.2"))
+    MAX_TOKENS: int = int(os.getenv("MAX_TOKENS", "2048"))
+    EMBEDDING_MODEL: str = os.getenv("EMBEDDING_MODEL", "models/embedding-001")
+
+    # =============================================================================
+    # VECTOR STORE CONFIGURATION
+    # =============================================================================
+    CHROMA_PERSIST_DIR: str = os.getenv("CHROMA_PERSIST_DIRECTORY", "./data/chroma_db")
+
+    # =============================================================================
+    # STORAGE CONFIGURATION
+    # =============================================================================
+    UPLOAD_DIR: str = os.getenv("UPLOAD_DIR", "./uploads")
+    DATA_DIR: str = os.getenv("DATA_DIR", "./data")
+    DATABASE_URL: str = os.getenv("DATABASE_URL", "sqlite:///./data/lega.db")
+
+    # =============================================================================
+    # PERFORMANCE SETTINGS
+    # =============================================================================
+    MAX_CONCURRENT_UPLOADS: int = int(os.getenv("MAX_CONCURRENT_UPLOADS", "5"))
+    DOCUMENT_PROCESSING_TIMEOUT: int = int(
+        os.getenv("DOCUMENT_PROCESSING_TIMEOUT", "300")
+    )
+    ENABLE_CACHE: bool = os.getenv("ENABLE_CACHE", "True").lower() == "true"
+    CACHE_TTL_SECONDS: int = int(os.getenv("CACHE_TTL_SECONDS", "3600"))
+
+    # =============================================================================
+    # FEATURE FLAGS
+    # =============================================================================
+    ENABLE_DOCUMENT_LIBRARY: bool = (
+        os.getenv("ENABLE_DOCUMENT_LIBRARY", "True").lower() == "true"
+    )
+    ENABLE_QA_ASSISTANT: bool = (
+        os.getenv("ENABLE_QA_ASSISTANT", "True").lower() == "true"
+    )
+    ENABLE_MARKET_COMPARISON: bool = (
+        os.getenv("ENABLE_MARKET_COMPARISON", "True").lower() == "true"
+    )
+    ENABLE_TIMELINE_TRACKER: bool = (
+        os.getenv("ENABLE_TIMELINE_TRACKER", "True").lower() == "true"
+    )
+    ENABLE_EXPORT_FEATURES: bool = (
+        os.getenv("ENABLE_EXPORT_FEATURES", "True").lower() == "true"
+    )
+
+    # =============================================================================
+    # ANALYTICS & MONITORING
+    # =============================================================================
+    ENABLE_ANALYTICS: bool = os.getenv("ENABLE_ANALYTICS", "False").lower() == "true"
+    ANALYTICS_API_KEY: str = os.getenv("ANALYTICS_API_KEY", "")
+    ENABLE_ERROR_TRACKING: bool = (
+        os.getenv("ENABLE_ERROR_TRACKING", "False").lower() == "true"
+    )
+    SENTRY_DSN: str = os.getenv("SENTRY_DSN", "")
+
+    # =============================================================================
+    # REGIONAL SETTINGS
+    # =============================================================================
+    DEFAULT_REGION: str = os.getenv("DEFAULT_REGION", "India")
+    DEFAULT_CURRENCY: str = os.getenv("DEFAULT_CURRENCY", "INR")
+    TIMEZONE: str = os.getenv("TIMEZONE", "Asia/Kolkata")
+
+    # =============================================================================
+    # ADVANCED AI SETTINGS
+    # =============================================================================
+    RISK_SENSITIVITY: int = int(os.getenv("RISK_SENSITIVITY", "3"))
+    SIMPLIFICATION_LEVEL: str = os.getenv("SIMPLIFICATION_LEVEL", "intermediate")
+    MAX_RISK_FACTORS: int = int(os.getenv("MAX_RISK_FACTORS", "10"))
+
+    # =============================================================================
+    # API RATE LIMITING
+    # =============================================================================
+    API_REQUESTS_PER_MINUTE: int = int(os.getenv("API_REQUESTS_PER_MINUTE", "60"))
+    API_REQUESTS_PER_DAY: int = int(os.getenv("API_REQUESTS_PER_DAY", "1000"))
+
+    # =============================================================================
+    # BACKUP & MAINTENANCE
+    # =============================================================================
+    ENABLE_AUTO_BACKUP: bool = (
+        os.getenv("ENABLE_AUTO_BACKUP", "False").lower() == "true"
+    )
+    BACKUP_INTERVAL_HOURS: int = int(os.getenv("BACKUP_INTERVAL_HOURS", "24"))
+    BACKUP_RETENTION_DAYS: int = int(os.getenv("BACKUP_RETENTION_DAYS", "30"))
+    AUTO_CLEANUP_TEMP_FILES: bool = (
+        os.getenv("AUTO_CLEANUP_TEMP_FILES", "True").lower() == "true"
+    )
+    CLEANUP_INTERVAL_HOURS: int = int(os.getenv("CLEANUP_INTERVAL_HOURS", "6"))
+
+    @classmethod
+    def validate_config(cls) -> bool:
+        """Validate that required configuration is present."""
+        required_vars = ["GOOGLE_API_KEY"]
+
+        missing_vars = []
+        for var in required_vars:
+            value = getattr(cls, var, "")
+            if not value or value == "your_google_ai_api_key_here":
+                missing_vars.append(var)
+
+        if missing_vars:
+            print(
+                f"⚠️  Missing required environment variables: {', '.join(missing_vars)}"
+            )
+            print("📝 Please update your .env file with valid values")
+            return False
+
+        return True
+
+    @classmethod
+    def get_config_summary(cls) -> dict:
+        """Get a summary of current configuration for debugging."""
+        return {
+            "api_configured": bool(
+                cls.GOOGLE_API_KEY
+                and cls.GOOGLE_API_KEY != "your_google_ai_api_key_here"
+            ),
+            "debug_mode": cls.DEBUG,
+            "features_enabled": {
+                "document_library": cls.ENABLE_DOCUMENT_LIBRARY,
+                "qa_assistant": cls.ENABLE_QA_ASSISTANT,
+                "market_comparison": cls.ENABLE_MARKET_COMPARISON,
+                "timeline_tracker": cls.ENABLE_TIMELINE_TRACKER,
+                "export_features": cls.ENABLE_EXPORT_FEATURES,
+            },
+            "supported_file_types": cls.SUPPORTED_FILE_TYPES,
+            "max_file_size_mb": cls.MAX_FILE_SIZE_MB,
+            "risk_sensitivity": cls.RISK_SENSITIVITY,
+            "region": cls.DEFAULT_REGION,
+            "currency": cls.DEFAULT_CURRENCY,
+        }
+
+
+# Create singleton instance
+config = Config()

src/utils/helpers.py

@@ -0,0 +1,172 @@
+import hashlib
+import os
+import uuid
+from datetime import datetime, timedelta
+from typing import List, Dict, Any, Optional
+import re
+
+
+def generate_document_id() -> str:
+    """Generate a unique document ID."""
+    return str(uuid.uuid4())
+
+
+def generate_session_id() -> str:
+    """Generate a unique session ID."""
+    return str(uuid.uuid4())
+
+
+def calculate_file_hash(file_content: bytes) -> str:
+    """Calculate SHA-256 hash of file content."""
+    return hashlib.sha256(file_content).hexdigest()
+
+
+def sanitize_filename(filename: str) -> str:
+    """Sanitize filename for safe storage."""
+    # Remove or replace dangerous characters
+    sanitized = re.sub(r"[^\w\-_\.]", "_", filename)
+    # Ensure it's not too long
+    if len(sanitized) > 255:
+        name, ext = os.path.splitext(sanitized)
+        sanitized = name[: 255 - len(ext)] + ext
+    return sanitized
+
+
+def format_file_size(size_bytes: int) -> str:
+    """Format file size in human readable format."""
+    if size_bytes == 0:
+        return "0 B"
+
+    size_names = ["B", "KB", "MB", "GB"]
+    i = 0
+    while size_bytes >= 1024 and i < len(size_names) - 1:
+        size_bytes /= 1024.0
+        i += 1
+
+    return f"{size_bytes:.1f} {size_names[i]}"
+
+
+def extract_key_dates(text: str) -> List[Dict[str, Any]]:
+    """Extract dates and deadlines from text."""
+    date_patterns = [
+        r"\b\d{1,2}/\d{1,2}/\d{4}\b",  # MM/DD/YYYY
+        r"\b\d{1,2}-\d{1,2}-\d{4}\b",  # MM-DD-YYYY
+        r"\b\d{4}-\d{1,2}-\d{1,2}\b",  # YYYY-MM-DD
+        r"\b(?:January|February|March|April|May|June|July|August|September|October|November|December)\s+\d{1,2},?\s+\d{4}\b",
+    ]
+
+    dates = []
+    for pattern in date_patterns:
+        matches = re.finditer(pattern, text, re.IGNORECASE)
+        for match in matches:
+            dates.append(
+                {
+                    "date": match.group(),
+                    "position": match.start(),
+                    "context": text[max(0, match.start() - 50) : match.end() + 50],
+                }
+            )
+
+    return dates
+
+
+def extract_financial_terms(text: str) -> Dict[str, Any]:
+    """Extract financial information from text."""
+    financial_info = {}
+
+    # Extract monetary amounts (Indian Rupees and other currencies)
+    money_patterns = [
+        r"₹[\d,]+(?:\.\d{2})?",  # Indian Rupees
+        r"Rs\.?\s*[\d,]+(?:\.\d{2})?",  # Rs. format
+        r"\$[\d,]+(?:\.\d{2})?",  # USD
+    ]
+
+    amounts = []
+    for pattern in money_patterns:
+        amounts.extend(re.findall(pattern, text))
+
+    if amounts:
+        financial_info["amounts"] = amounts
+
+    # Extract percentages
+    percentage_pattern = r"\d+(?:\.\d+)?%"
+    percentages = re.findall(percentage_pattern, text)
+    if percentages:
+        financial_info["percentages"] = percentages
+
+    # Extract interest rates
+    interest_pattern = (
+        r"(?:interest rate|APR|annual percentage rate).*?(\d+(?:\.\d+)?%)"
+    )
+    interest_matches = re.findall(interest_pattern, text, re.IGNORECASE)
+    if interest_matches:
+        financial_info["interest_rates"] = interest_matches
+
+    return financial_info
+
+
+def calculate_risk_score(risk_factors: List[Dict[str, Any]]) -> int:
+    """Calculate overall risk score from individual risk factors."""
+    if not risk_factors:
+        return 0
+
+    risk_weights = {"critical": 25, "high": 15, "medium": 8, "low": 3}
+
+    total_score = 0
+    for factor in risk_factors:
+        severity = factor.get("severity", "low").lower()
+        total_score += risk_weights.get(severity, 0)
+
+    # Cap at 100
+    return min(total_score, 100)
+
+
+def get_risk_color(risk_score: int) -> str:
+    """Get color code based on risk score."""
+    if risk_score >= 75:
+        return "#FF4444"  # Red
+    elif risk_score >= 50:
+        return "#FF8800"  # Orange
+    elif risk_score >= 25:
+        return "#FFCC00"  # Yellow
+    else:
+        return "#44AA44"  # Green
+
+
+def chunk_text(text: str, chunk_size: int = 1000, overlap: int = 100) -> List[str]:
+    """Split text into overlapping chunks for processing."""
+    chunks = []
+    start = 0
+
+    while start < len(text):
+        end = start + chunk_size
+        chunk = text[start:end]
+
+        # Try to break at sentence boundary
+        if end < len(text):
+            last_period = chunk.rfind(".")
+            if last_period > chunk_size // 2:
+                chunk = chunk[: last_period + 1]
+                end = start + last_period + 1
+
+        chunks.append(chunk)
+        start = end - overlap
+
+    return chunks
+
+
+def format_timestamp(timestamp: datetime) -> str:
+    """Format timestamp for display."""
+    now = datetime.now()
+    diff = now - timestamp
+
+    if diff.days > 0:
+        return f"{diff.days} days ago"
+    elif diff.seconds > 3600:
+        hours = diff.seconds // 3600
+        return f"{hours} hours ago"
+    elif diff.seconds > 60:
+        minutes = diff.seconds // 60
+        return f"{minutes} minutes ago"
+    else:
+        return "Just now"

src/utils/logger.py

@@ -0,0 +1,58 @@
+import logging
+import os
+from datetime import datetime
+from src.utils.config import config
+
+
+def setup_logging():
+    """Set up logging configuration."""
+    # Create data directory if it doesn't exist
+    os.makedirs(os.path.dirname(config.LOG_FILE), exist_ok=True)
+
+    # Configure logging
+    logging.basicConfig(
+        level=getattr(logging, config.LOG_LEVEL),
+        format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
+        handlers=[logging.FileHandler(config.LOG_FILE), logging.StreamHandler()],
+    )
+
+    return logging.getLogger(__name__)
+
+
+def log_document_upload(filename: str, file_size: int) -> None:
+    """Log document upload event."""
+    logger = logging.getLogger(__name__)
+    logger.info(f"Document uploaded: {filename} ({file_size} bytes)")
+
+
+def log_analysis_start(document_id: str) -> None:
+    """Log analysis start event."""
+    logger = logging.getLogger(__name__)
+    logger.info(f"Starting analysis for document: {document_id}")
+
+
+def log_analysis_complete(document_id: str, processing_time: float) -> None:
+    """Log analysis completion event."""
+    logger = logging.getLogger(__name__)
+    logger.info(
+        f"Analysis completed for document: {document_id} in {processing_time:.2f}s"
+    )
+
+
+def log_error(error_message: str, document_id: str = None) -> None:
+    """Log error event."""
+    logger = logging.getLogger(__name__)
+    if document_id:
+        logger.error(f"Error processing document {document_id}: {error_message}")
+    else:
+        logger.error(f"Application error: {error_message}")
+
+
+def log_qa_interaction(document_id: str, question: str) -> None:
+    """Log Q&A interaction."""
+    logger = logging.getLogger(__name__)
+    logger.info(f"Q&A interaction for document {document_id}: {question[:100]}...")
+
+
+# Initialize logging when module is imported
+setup_logging()

start.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+
+# Hugging Face Spaces startup script for Lega.AI
+
+echo "🚀 Starting Lega.AI on Hugging Face Spaces..."
+
+# Create necessary directories if they don't exist
+mkdir -p data/chroma_db
+mkdir -p uploads
+mkdir -p .streamlit
+
+# Set default environment variables for Hugging Face deployment
+export STREAMLIT_SERVER_PORT=${PORT:-7860}
+export STREAMLIT_SERVER_ADDRESS="0.0.0.0"
+export DEBUG=False
+export LOG_LEVEL=INFO
+export STREAMLIT_CONFIG_DIR=/app/.streamlit
+export XDG_CONFIG_HOME=/app
+
+# Check if GOOGLE_API_KEY is set
+if [ -z "$GOOGLE_API_KEY" ]; then
+    echo "⚠️  WARNING: GOOGLE_API_KEY environment variable is not set!"
+    echo "Please set it in your Hugging Face Space settings for the app to work properly."
+fi
+
+# Start the Streamlit application
+echo "🌐 Starting Streamlit on port $STREAMLIT_SERVER_PORT..."
+exec streamlit run main.py \
+    --server.port=$STREAMLIT_SERVER_PORT \
+    --server.address=$STREAMLIT_SERVER_ADDRESS \
+    --server.headless=true \
+    --server.fileWatcherType=none \
+    --server.enableCORS=false \
+    --server.enableXsrfProtection=false