🧠 Universal Document RAG System using Groq API

A high-performance Retrieval-Augmented Generation (RAG) system that enables users to ask natural language questions over their own documents. The application retrieves the most relevant information using semantic vector search and generates accurate, context-aware responses using Groq-hosted Llama 3.1 models.

Designed with a modular architecture, this project serves as a foundation for building AI-powered document assistants, enterprise knowledge bases, research assistants, and intelligent search systems.

---

✨ Features

• 📄 Supports multiple document formats (PDF, TXT, JSON)
• 🔍 Semantic search using Sentence Transformers
• ⚡ Ultra-fast inference powered by Groq API
• 🧠 Retrieval-Augmented Generation (RAG) for grounded responses
• 📚 Automatic document chunking and embedding generation
• 💾 High-speed similarity search using FAISS
• 🔐 Secure API key management with .env
• 🏗️ Modular and scalable project architecture
• 💻 Interactive CLI chat interface
• 🚀 Easy to extend with additional document formats

---

🏗️ Architecture Overview

flowchart TD
    A([User Query])
    B[Generate Query Embedding]
    C[FAISS Vector Search]
    D[Retrieve Top-K Relevant Chunks]
    E[Construct Prompt with Retrieved Context]
    F[Groq API

Llama 3.1]
    G([Context-Aware Response])

    A --> B
    B --> C
    C --> D
    D --> E
    E --> F
    F --> G

---

🔄 RAG Workflow

flowchart TD

    A[Load Documents]
    B[Split into Chunks]
    C[Generate Embeddings]
    D[Build FAISS Index]
    E[Store Vector Database]
    F[User Query]
    G[Generate Query Embedding]
    H[Semantic Search]
    I[Retrieve Top-K Chunks]
    J[Prompt Construction]
    K[Groq Llama 3.1]
    L[Final Answer]

    A --> B
    B --> C
    C --> D
    D --> E
    E --> F
    F --> G
    G --> H
    H --> I
    I --> J
    J --> K
    K --> L

---

🛠️ Tech Stack

Component	Technology	Purpose
Language	Python	Core development
Embeddings	Sentence Transformers	Semantic vector embeddings
Vector Database	FAISS	Fast similarity search
Large Language Model	Llama 3.1	Context-aware response generation
LLM Inference	Groq API	Low-latency inference
Environment Variables	python-dotenv	Secure API key handling

---

📁 Project Structure

doc_RAG/
│
├── core/
│   ├── embedder.py          # Embedding model
│   ├── vector_db.py         # FAISS index operations
│   ├── retriever.py         # Semantic retrieval
│   └── groq_llm.py          # Groq API integration
│
├── loaders/
│   ├── __init__.py
│   ├── json_loader.py
│   ├── pdf_loader.py
│   └── text_loader.py
│
├── data/
│   ├── colleges.json
│   ├── college_from_js.json
│   ├── sample.txt
│   └── document.pdf
│
├── build_index.py           # Builds the vector database
├── rag_chat.py              # Interactive chatbot
├── requirements.txt
├── .env
└── README.md

---

⚙️ Installation

1️⃣ Clone the Repository

git clone https://github.com/your-username/universal-document-rag.git

cd universal-document-rag

---

2️⃣ Install Dependencies

pip install -r requirements.txt

---

3️⃣ Configure Environment Variables

Create a .env file in the project root.

GROQ_API_KEY=your_groq_api_key_here

⚠️ Never commit your .env file to GitHub.

---

4️⃣ Build the Vector Database

python build_index.py

Expected Output:

Loading documents...
Generating embeddings...
Creating FAISS index...
Vector database built successfully.

---

5️⃣ Launch the Chat Interface

python rag_chat.py

---

💬 Example Usage

Query

Where is Ariyalur Engineering College located?

Response

Ariyalur Engineering College is located at:

NH-227, Trichy–Chidambaram Highway,
Karuppur-Senapathy Post,
Ariyalur District,
Tamil Nadu.

---

🎯 Applications

• 📚 Document Question Answering
• 🏢 Enterprise Knowledge Bases
• 🎓 Educational Assistants
• 📑 Research Paper Search
• ⚖️ Legal Document Analysis
• 🏥 Healthcare Knowledge Systems
• 🤖 AI-powered Internal Chatbots

---

🚀 Future Enhancements

• 🌐 Streamlit or FastAPI Web Interface
• 🗂️ Support for DOCX, CSV, and HTML documents
• 📤 Drag-and-drop document upload
• 💬 Multi-turn conversation memory
• 🔄 Incremental vector indexing
• 📖 Source citations with confidence scores
• 🖼️ OCR support for scanned PDFs
• 🌍 Multilingual document support
• ☁️ Integration with Pinecone, ChromaDB, and Milvus

---

🤝 Contributing

Contributions, issues, and feature requests are welcome!

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

---

📜 License

This project is licensed under the MIT License.

---

⭐ Support

If you found this project useful, please consider giving it a ⭐ Star on GitHub.

Your support helps improve the project and encourages future development.