Hugging Face's logo

Spaces:
ratulsur
/
BI_ANALYTICS
Configuration error

App Files Community
BI_ANALYTICS / README.md
ratulsur's picture
ratulsur
Upload 8 files
e51a81b verified
preview code
|
raw
history blame contribute delete
10.3 kB

BI Storyteller - Python Standard Library Edition

A comprehensive marketing analysis automation platform built entirely with Python's standard library, designed to work in WebContainer environments.

🎯 Features

Complete 12-Module Workflow

  1. 🔑 API Key Setup - Secure Groq API integration
  2. 📝 Variable Extraction - AI-powered variable identification
  3. 📋 Questionnaire Generator - Dynamic survey creation
  4. 🔢 Data Generation - Realistic sample data creation
  5. 🧹 Data Cleaning - Comprehensive preprocessing
  6. 📊 EDA Module - Statistical analysis with AI insights
  7. 🤖 Predictive Analytics - Machine learning simulation
  8. 📈 Trend Analysis - Time series analysis and forecasting
  9. 💭 Sentiment Analysis - Customer feedback analysis
  10. 🧪 A/B Testing - Statistical significance testing
  11. 💬 Chat with Data - Interactive AI-powered Q&A
  12. 📤 Export & Import - Data persistence and sharing

Key Capabilities

  • 🌐 Web Interface - Professional browser-based UI
  • 💻 CLI Interface - Command-line interaction option
  • 🤖 AI Integration - Groq LLM for intelligent insights
  • 📊 Statistical Analysis - Comprehensive data analysis
  • 🔄 Workflow Management - Sequential module progression
  • 💾 Data Persistence - CSV and JSON export/import
  • 🎨 Professional UI - Clean, modern web interface

🚀 Quick Start

Option 1: Web Interface (Recommended) 

python main.py

Then open your browser to http://localhost:8000

Option 2: Command Line Interface 

python cli_interface.py

📋 Requirements

Environment: Python 3.6+ (Standard Library Only)

  • No external dependencies required
  • Works in WebContainer, local environments, and cloud platforms
  • Compatible with restricted Python environments

Optional: Groq API key for AI-powered features

  • Get free API key at console.groq.com
  • Fallback functionality available without API key

🎮 Usage Guide

Web Interface Workflow

  1. 🔑 Set API Key

    • Enter your Groq API key for AI features
    • Skip this step to use fallback functionality

  2. 📝 Extract Variables

    • Describe your business problem
    • AI extracts relevant marketing variables
    • Review and proceed to next step

  3. 📋 Generate Questionnaire

    • Automatically creates survey questions
    • Based on extracted variables
    • Mix of multiple choice and descriptive questions

  4. 🔢 Generate Sample Data

    • Creates realistic sample dataset
    • Configurable sample size (100-10,000 records)
    • Based on your specific variables

  5. 🧹 Clean Data

    • Handles missing values and outliers
    • Removes duplicates
    • Provides cleaning statistics

  6. 📊 Perform EDA

    • Statistical analysis and correlations
    • AI-generated insights
    • Distribution analysis

  7. 🤖 Train Predictive Model

    • Multiple algorithm options
    • Performance metrics simulation
    • Feature importance analysis

  8. 📈 Analyze Trends

    • Time series analysis
    • Seasonality detection
    • Revenue forecasting

  9. 💭 Analyze Sentiment

    • Customer feedback analysis
    • Sentiment distribution
    • Actionable recommendations

  10. 🧪 Run A/B Test

    • Statistical significance testing
    • Conversion rate analysis
    • Winner determination

  11. 💬 Chat with Data

    • Interactive Q&A about your analysis
    • AI-powered insights
    • Context-aware responses

  12. 📤 Export Results

    • Download complete analysis as JSON
    • Save data as CSV files
    • Share results with stakeholders

Command Line Interface

The CLI provides the same functionality through an interactive menu system:

python cli_interface.py

Navigate through numbered options (1-15) to complete your analysis workflow.

🔧 Technical Architecture

Core Components

main.py - Core BIStoryteller class with all analysis methods

  • Variable extraction with AI integration
  • Data generation and cleaning algorithms
  • Statistical analysis and correlation calculations
  • Predictive modeling simulation
  • Trend analysis and forecasting
  • Sentiment analysis with rule-based fallbacks
  • A/B testing with statistical significance
  • Chat interface with contextual responses

web_interface.py - HTTP server with REST API

  • Professional HTML/CSS/JavaScript interface
  • RESTful API endpoints for all modules
  • Real-time status updates
  • Responsive design for all devices

cli_interface.py - Command-line interface

  • Interactive menu system
  • Formatted output displays
  • Progress tracking
  • User-friendly prompts

Data Flow

  1. Input → Business problem description
  2. Processing → AI variable extraction
  3. Generation → Sample data creation
  4. Cleaning → Data preprocessing
  5. Analysis → Statistical insights
  6. Modeling → Predictive analytics
  7. Visualization → Trend and sentiment analysis
  8. Testing → A/B test evaluation
  9. Interaction → Chat-based exploration
  10. Export → Results and data export

🌟 Key Features

AI-Powered Analysis

  • Groq LLM Integration for intelligent variable extraction
  • Contextual Questionnaire Generation based on business problems
  • Smart Insights Generation from statistical analysis
  • Interactive Chat with your data and results

Statistical Capabilities

  • Correlation Analysis with Pearson coefficients
  • Distribution Analysis with descriptive statistics
  • Trend Detection with significance testing
  • A/B Testing with proper statistical methods
  • Forecasting with confidence intervals

Professional Interface

  • Modern Web UI with responsive design
  • Real-time Updates and progress indicators
  • Error Handling with user-friendly messages
  • Data Visualization through statistical summaries
  • Export Capabilities for sharing and persistence

Robust Architecture

  • Modular Design with clear separation of concerns
  • Error Handling with graceful fallbacks
  • Data Validation at every step
  • Memory Efficient processing
  • Cross-platform compatibility

📊 Sample Workflow 

# Initialize BI Storyteller
bi = BIStoryteller()

# Set API key
bi.set_groq_api_key("your_groq_api_key")

# Extract variables
variables = bi.extract_variables("We want to improve customer retention and increase purchase frequency")

# Generate questionnaire
questionnaire = bi.generate_questionnaire(variables, business_problem)

# Generate and clean data
sample_data = bi.generate_sample_data(variables, 1000)
cleaned_data, cleaning_results = bi.clean_data(sample_data)

# Perform analysis
eda_results = bi.perform_eda(cleaned_data)
model_results = bi.train_predictive_model(cleaned_data, "Random Forest")
trend_results = bi.analyze_trends(cleaned_data, "Monthly")

# Interactive analysis
response = bi.chat_with_data("What are the key factors driving customer satisfaction?")

# Export results
bi.export_results("my_analysis.json")

🛠️ Customization

Adding New Analysis Methods

Extend the BIStoryteller class with new methods:

def custom_analysis(self, data, parameters):
    """Add your custom analysis logic"""
    # Your analysis code here
    return results

Modifying the Web Interface

Edit the HTML template in web_interface.py to customize:

  • UI styling and layout
  • Form fields and options
  • Result display formats
  • Additional functionality

Extending API Endpoints

Add new endpoints in the do_POST method:

elif self.path == '/api/custom_endpoint':
    # Handle custom functionality
    result = self.bi.custom_analysis(data)
    self._send_json_response({'results': result})

🔒 Security & Privacy

  • API keys stored in memory only (not persisted)
  • No external data transmission except to Groq API
  • Local data processing - your data stays on your machine
  • No tracking or analytics - completely private

🚀 Deployment Options

Local Development 

python main.py  # Web interface on localhost:8000
python cli_interface.py  # Command line interface

Cloud Deployment

The application can be deployed to any Python-supporting platform:

  • Heroku, Railway, Render
  • Google Cloud Run, AWS Lambda
  • Any VPS with Python support

Docker Deployment 

FROM python:3.9-slim
COPY . /app
WORKDIR /app
EXPOSE 8000
CMD ["python", "main.py"]

🎯 Use Cases

Marketing Teams

  • Customer segmentation analysis
  • Campaign performance optimization
  • A/B testing for marketing materials
  • Customer satisfaction tracking

Data Analysts

  • Automated EDA workflows
  • Predictive modeling pipelines
  • Trend analysis and forecasting
  • Statistical significance testing

Business Consultants

  • Client data analysis automation
  • Professional reporting generation
  • Interactive data exploration
  • Stakeholder presentation preparation

Research Teams

  • Survey design and analysis
  • Statistical hypothesis testing
  • Data cleaning and preprocessing
  • Collaborative analysis workflows

🤝 Support

Getting Help

  • Check error messages in terminal/browser console
  • Review the workflow sequence (modules must be completed in order)
  • Verify API key is set correctly for AI features

Troubleshooting

  • Port already in use: Change port in start_web_server(port=8001)
  • API errors: Check Groq API key validity and internet connection
  • Data issues: Ensure previous modules are completed before proceeding

Feature Requests

The application is designed to be comprehensive and extensible. You can:

  • Add custom analysis methods to the core class
  • Extend the web interface with new modules
  • Integrate additional AI providers
  • Customize the statistical analysis methods

📄 License

This project is provided as-is for educational and commercial use.

🚀 Ready to automate your marketing analysis? Start with the web interface or CLI and transform your business problems into actionable insights!

Quick Commands: 

# Start web interface
python main.py

# Start CLI interface  
python cli_interface.py

# View this help
python -c "import main; help(main.BIStoryteller)"