🏥 GENAI Smart Medical Chatbot

An AI-powered medical assistant that answers health questions using Retrieval-Augmented Generation (RAG), with a full chat history sidebar, conversation memory, and a modern UI.

Features • Tech Stack • Installation • API Endpoints • Deployment

---

📸 Screenshots

Chat Interface	Chat History Sidebar

---

✨ Features

• 🤖 AI Medical Q&A — Ask any medical question and get specific, actionable answers
• 💬 Chat History Sidebar — Browse and resume previous conversations (like Claude / ChatGPT)
• ➕ New Chat Button — Start fresh conversations anytime; previous chats are auto-saved
• 🧠 Conversation Memory — Bot remembers context within each session for natural follow-ups
• 🔍 RAG Pipeline — Retrieves relevant medical knowledge from Pinecone before generating answers
• 📝 Markdown Rendering — Responses display with bold text, numbered lists, and headers
• 🗄️ Persistent History — All chats stored in local SQLite database across restarts
• 📱 Responsive UI — Collapsible sidebar works on desktop and mobile
• 🚀 Cloud Inference — No local GPU needed; uses HuggingFace Inference API

---

🛠️ Tech Stack

Layer	Technology
Backend	Python 3.10+, Flask 3.0
Containerization	Docker (HuggingFace Spaces ready, port 7860)
LLM	Llama 3.1 8B Instruct via HuggingFace Inference API
RAG Framework	LangChain 0.2
Vector Database	Pinecone (index: medical)
Embeddings	sentence-transformers/all-MiniLM-L6-v2
Chat Storage	SQLite (local, zero-config)
Frontend	HTML5, CSS3, JavaScript, jQuery, marked.js, DOMPurify

---

📂 Project Structure

GENAI-Smart-Medical-Chatbot/
├── Dockerfile               # Docker image — used for HuggingFace Spaces deployment
├── app.py                   # Flask app — routes, RAG chain, SQLite logic
├── store_index.py           # Script to embed documents and populate Pinecone
├── requirements.txt         # Python dependencies
├── setup.py                 # Package setup
├── .env                     # Environment variables (not committed)
├── chat_history.db          # SQLite database (auto-created at runtime)
│
├── data/
│   └── blood_sugar_guide.txt  # Curated medical knowledge (blood sugar management)
│
├── src/
│   ├── helper.py            # PDF loading and text splitting utilities
│   └── prompt.py            # Legacy prompt template
│
├── model/
│   └── instructions.txt     # Model download instructions
│
├── templates/
│   └── chat.html            # Full frontend (sidebar + chat UI)
│
├── static/                  # Static assets (CSS, images)
└── research/                # Notebooks and experiments

---

⚙️ Installation & Setup

Prerequisites

• Python 3.9+
• A HuggingFace account (free) with an API token
• A Pinecone account (free tier) with an API key

1. Clone the repository

git clone https://github.com/agrawal-2005/GENAI-Smart-Medical-Chatbot.git
cd GENAI-Smart-Medical-Chatbot

2. Create and activate a virtual environment

python -m venv venv

# macOS / Linux
source venv/bin/activate

# Windows
venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. Create your .env file

cp .env.example .env   # or create manually

Add the following to .env:

HUGGINGFACEHUB_API_TOKEN=your_huggingface_token_here
PINECONE_API_KEY=your_pinecone_api_key_here

5. Populate the Pinecone index (first time only)

Place any medical PDF files in the data/ folder, then run:

python store_index.py

The curated data/blood_sugar_guide.txt is already indexed. Only re-run this when adding new documents.

6. Run the application

python app.py

Open your browser and visit:

http://localhost:7860

Or run with Docker

docker build -t medical-chatbot .
docker run -p 7860:7860 \
  -e HUGGINGFACEHUB_API_TOKEN=your_token \
  -e PINECONE_API_KEY=your_key \
  medical-chatbot

---

🔑 Environment Variables

Variable	Description	Where to get it
HUGGINGFACEHUB_API_TOKEN	HuggingFace API token for Llama 3.1 inference	huggingface.co/settings/tokens
PINECONE_API_KEY	Pinecone vector database API key	app.pinecone.io

⚠️ Never commit your .env file. It is listed in .gitignore.

---

🔌 API Endpoints

Method	Endpoint	Description
GET	/	Main chat interface
POST	/get	Send a message; returns AI response and conversation_id
GET	/api/conversations	List all saved conversations
POST	/api/conversations	Create a new blank conversation
GET	/api/conversations/<id>	Get a conversation with its full message history
DELETE	/api/conversations/<id>	Delete a conversation and all its messages

Example: Send a message

curl -X POST http://localhost:7860/get \
  -F "msg=What is my target blood sugar range?" \
  -F "conversation_id=optional-uuid-here"

{
  "response": "For most adults with diabetes, target blood sugar levels are:\n- **Before meals:** 80–130 mg/dL\n- **After meals (2hr):** < 180 mg/dL",
  "conversation_id": "abc123-..."
}

---

🧠 How It Works

User Message
     │
     ▼
┌─────────────────────┐
│  Pinecone Retriever │  ← Finds top-3 relevant medical document chunks
└─────────────────────┘
     │ context
     ▼
┌─────────────────────┐
│   Prompt Builder    │  ← Injects context + last 3 conversation turns
└─────────────────────┘
     │ formatted prompt
     ▼
┌─────────────────────┐
│  Llama 3.1 8B via  │  ← HuggingFace Inference API (no local GPU needed)
│   HF Router API    │
└─────────────────────┘
     │ answer
     ▼
┌─────────────────────┐
│  SQLite + Response  │  ← Saves to DB, renders markdown, returns to UI
└─────────────────────┘

---

🚀 Deployment

This app requires no local model download — inference runs entirely in the cloud via the HuggingFace API.

⭐ HuggingFace Spaces (Recommended)

The project includes a Dockerfile configured for HF Spaces (port 7860).

1. Go to huggingface.co/new-space
2. Choose Docker as the Space SDK
3. Link your GitHub repo or clone the Space and push manually:

   git remote add space https://huggingface.co/spaces/YOUR_USERNAME/YOUR_SPACE_NAME
   git push space main

4. Add secrets in Settings → Variables and Secrets:

   Secret	Value
   HUGGINGFACEHUB_API_TOKEN	Your HF token
   PINECONE_API_KEY	Your Pinecone key

5. The Space will build the Docker image and go live automatically at:

   https://huggingface.co/spaces/YOUR_USERNAME/YOUR_SPACE_NAME
   

⚠️ SQLite chat history will reset on each Space restart. For persistent history, replace SQLite with a hosted database (PostgreSQL, MongoDB Atlas, etc.)

🐳 Docker (Self-hosted)

docker build -t medical-chatbot .
docker run -p 7860:7860 \
  -e HUGGINGFACEHUB_API_TOKEN=your_token \
  -e PINECONE_API_KEY=your_key \
  medical-chatbot

Render / Railway

1. Connect your GitHub repo
2. Set start command: python app.py
3. Add HUGGINGFACEHUB_API_TOKEN and PINECONE_API_KEY as environment variables
4. Set port to 7860

---

🤝 Contributing

Contributions are welcome!

1. Fork the repository
2. Create a feature branch: git checkout -b feature/your-feature-name
3. Commit your changes: git commit -m "Add your feature"
4. Push to the branch: git push origin feature/your-feature-name
5. Open a Pull Request

Please make sure your code runs without errors before submitting.

---

📄 License

This project is licensed under the MIT License — see the LICENSE file for details.

---

👨‍💻 Author

Prashant Agrawal

---

⭐ If you find this project useful, please give it a star! ⭐