RAG Multi-Agent Q&A Assistant

This project is a lightweight retrieval-augmented question-answering system that uses an agentic workflow to decide how to handle user queries. It can:

• Retrieve and answer from uploaded knowledge documents (RAG)
• Route to a calculator or dictionary tool when needed

🚀 Features

• RAG with FAISS vector DB
• Simple agent routing logic (calculator/dictionary vs retrieval)
• Streamlit UI with decision logging and results

🛠️ How to Run

1. Clone the repo
2. Install dependencies:

   pip install -r requirements.txt

3. Set your OpenAI API key:

   export OPENAI_API_KEY=your-key-here

4. Run the app:

   streamlit run app.py

Sample Data

Place your .txt documents inside the data/ folder.

🧱 Technologies Used

• LangChain
• FAISS
• Streamlit
• OpenAI (or HuggingFace)

📸 Screenshots

Deployments

Streamlit Link:- https://rag-multi-agent-q-a-assistant-okoegbtpytatmespksfomu.streamlit.app/

Contact

Author:- Sd-space Contact:-ssrvdd@gmail.com