README.md

---
title: Sagar's Personal Assistant
emoji: 🌟
colorFrom: purple
colorTo: blue
sdk: streamlit
sdk_version: "1.41.0"
app_file: app.py
pinned: false
---

# Sagar's Personal Assistant 🌟

A friendly, personal RAG-based chatbot that answers questions about Sagar using AI and documents (`resume.pdf`, `myself.txt`).

## Features
- 🤖 Powered by Google AI Studio Gemini 2.5 Flash
- 💬 Friendly, warm personality with emoji responses
- 📚 RAG (Retrieval-Augmented Generation) for accurate answers
- 🎨 Clean Streamlit web interface

## Setup

1.  **Install Dependencies**:
    ```bash
    pip install -r requirements.txt
    ```

2.  **Environment Variables**:
    Ensure `.env` exists with your `GOOGLE_API_KEY`.

## Usage

1.  **Ingest Data**:
    Process your PDFs and text files to create the vector database.
    ```bash
    python src/ingest.py
    ```
    *Run this whenever you add new files to the `data/` folder.*

2.  **Run Chatbot (Web Interface)**:
    Start the Streamlit web chat interface.
    ```bash
    streamlit run app.py
    ```

3.  **Run Chatbot (CLI)**:
    Start the command-line interface.
    ```bash
    python src/main.py
    ```

## Project Structure

- `data/`: Place your PDF and TXT files here.
- `app.py`: Streamlit web application.
- `src/rag.py`: Core RAG logic (Retrieval + Generation).
- `src/main.py`: Command-line interface.
- `src/ingest.py`: Script to load data and generate embeddings (FAISS).
- `src/vectorstore/`: Stores the generated FAISS index.