🧠 GenAI RAG Chatbot (LangChain + HuggingFace + FAISS)

A Retrieval-Augmented Generation (RAG) chatbot that answers questions only from provided documents, built using LangChain, Hugging Face models, FAISS vector search, and Streamlit.

---

🚀 Features

• 🔍 Semantic search using FAISS
• 🧠 Hugging Face LLM (FLAN-T5)
• 📄 Answers grounded in documents
• 🔐 Input safety filtering
• 📚 Shows source documents for transparency
• 🖥️ Simple Streamlit web UI

---

🚀 Live Demo

Check out the deployed app here: GenAI RAG Chatbot

---

🏗️ Project Architecture

User Question
     ↓
Safety Check
     ↓
Retriever (FAISS)
     ↓
Relevant Context
     ↓
LLM (Hugging Face)
     ↓
Answer + Source Documents

📁 Project Structure

rag_doc_assistant_genai/
│
├── scripts/
│   ├── embeddings.py        # Vector store (FAISS)
│   ├── rag_pipeline.py      # RAG chain logic
│   ├── safety.py            # Input safety checks
│   ├── evaluation.py        # RAG evaluation
│   └── finetune.py          # Fine-tuning experiments
│
├── data/
│   └── sample.txt           # Knowledge source
│
├── app_streamlit.py         # Streamlit UI
├── requirements.txt         # Dependencies
└── README.md

⚙️ Setup Instructions

1️⃣ Create & Activate Virtual Environment

# Create Virtual Environment
  python -m venv genai_env

# Activate it:
  # On Windows:
    genai_env\Scripts\activate
  # On Mac/Linux:
    source genai_env/bin/activate

2️⃣ Install Dependencies

pip install -r requirements.txt

📄 Add Your Data

Place your documents inside:

data/sample.txt  #The chatbot answers only from this file.

3️⃣ Run the Streamlit App

streamlit run app_streamlit.py

💡 Example

Question:

Where was the Kalinga War fought?

Answer:

The Kalinga War was fought at Dhauli, near present-day Bhubaneswar.

Source:

sample.txt 

⚙️ Tech Stack

• LLM: Hugging Face (FLAN-T5)
• Embeddings: Sentence Transformers
• Vector DB: FAISS
• Framework: LangChain
• Frontend: Streamlit
• Language: Python

👩‍💻 Author

Anil Kumar Nayak

✨ Software Developer | Python, AI & Streamlit Enthusiast

📧 anilnayak.tech@gmail.com

🏁 Future Enhancements

• PDF & Web-based RAG

• Multi-document support

• OpenAI / LLaMA integration

• Chat memory

• Docker deployment