📄 Document RAG Assistant

A powerful Document-based Retrieval Augmented Generation (RAG) system built with ChromaDB, Google's Gemini 2.0 Flash, and Streamlit. Upload PDF documents and ask questions based strictly on their content - no hallucinations, only factual answers from your document.

🚀 Features

• 📁 PDF Document Upload: Extract text from PDF files automatically
• 🔍 Smart Text Chunking: Intelligently splits documents for optimal retrieval
• 🎯 Semantic Search: Find relevant document sections using vector similarity
• 🤖 Anti-Hallucination AI: Responses based only on document content
• 📖 Source Attribution: See exactly which document sections were used
• 🗑️ Database Management: Clear and manage your document collection
• ⚡ Real-time Processing: Fast document processing and query responses

🛠️ Setup

Prerequisites

• Python 3.8+
• Gemini API key from Google AI Studio

Installation

1. Clone the repository:

git clone <your-repo-url>
cd "RAG DEMO"

2. Install required packages:

pip install -r requirements.txt

3. Set up environment variables: Create a .env file in the root directory:

GEMINI_API_KEY=your_gemini_api_key_here
CHROMA_API_KEY=your_chroma_api_key_here
CHROMA_TENANT=your_chroma_tenant_id
CHROMA_DATABASE=your_database_name

4. Run the application:

streamlit run app.py

📋 Usage Guide

1. Upload Documents

• Click "Choose a PDF file" to upload your document
• Wait for processing (text extraction + chunking)
• See confirmation with chunk count and preview

2. Ask Questions

• Type your question in the text input field
• Click "🔍 Get Answer" to search and generate response
• Review the AI answer and source document sections

3. Manage Documents

• Use the sidebar to see document chunk count
• Click "🗑️ Clear Database" to remove all documents
• Upload multiple documents for cross-document queries

🎯 Key Benefits

✅ No Hallucinations

• AI responds only based on uploaded document content
• Says "I don't know" when information isn't available
• No external knowledge or made-up information

✅ Source Transparency

• Shows relevant document sections used for answers
• Displays filename and chunk information
• Enables fact-checking and verification

✅ Smart Retrieval

• Uses semantic similarity for relevant content discovery
• Configurable chunk size (1000 chars) with overlap (200 chars)
• Retrieves top 5 most relevant sections per query

🏗️ Architecture

PDF Upload → Text Extraction → Text Chunking → Vector Embeddings → ChromaDB Storage
                                                                            ↓
User Query → Semantic Search → Relevant Chunks → Context Building → Gemini 2.0 → Response

🔧 Technical Components

• Streamlit: Web interface and user experience
• ChromaDB: Vector database for document storage
• Google Gemini 2.0 Flash: Large language model for response generation
• PyPDF2: PDF text extraction
• LangChain: Text splitting and chunking utilities

📁 Project Structure

RAG DEMO/
├── app.py                 # Main Streamlit application
├── requirements.txt       # Python dependencies
├── .env                  # Environment variables (not in git)
├── .gitignore           # Git ignore rules
└── README.md            # This file

⚙️ Configuration

Environment Variables

• GEMINI_API_KEY: Your Google AI Studio API key
• CHROMA_API_KEY: ChromaDB cloud API key (optional)
• CHROMA_TENANT: ChromaDB tenant ID (optional)
• CHROMA_DATABASE: ChromaDB database name (optional)

Chunking Parameters

• Chunk Size: 1000 characters (adjustable in code)
• Chunk Overlap: 200 characters (prevents information loss)
• Retrieval Count: 5 most relevant chunks per query

🔒 Security

• API keys stored in environment variables
• .env file excluded from version control
• No sensitive data in code repository
• Local ChromaDB instance for privacy

🚫 Limitations

• PDF Only: Currently supports PDF documents only
• Text-based: Cannot process images, tables, or complex layouts
• Memory Storage: Uses in-memory ChromaDB (data lost on restart)
• Single Session: No persistent user sessions or multi-user support

🛣️ Future Enhancements

• Support for Word documents, text files, and web pages
• Persistent ChromaDB storage
• Multi-user authentication and sessions
• Advanced document preprocessing (tables, images)
• Conversation history and follow-up questions
• Document comparison and analysis features

📄 License

This project is open source and available under the MIT License.

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Push to the branch
5. Create a Pull Request

📞 Support

If you encounter any issues or have questions:

• Check the troubleshooting section in the app
• Review error messages in the Streamlit interface
• Ensure your API keys are correctly configured
• Verify PDF files contain extractable text

---

Happy Document Querying! 🎉