🩺 RxAI – Medical RAG-Based Chatbot

RxAI is an intelligent medical question-answering system powered by Retrieval-Augmented Generation (RAG), Pinecone vector search, HuggingFace embeddings, and LangChain.
Users can ask questions about symptoms, cures, diseases, medicines, and precautions, and RxAI provides accurate responses backed by verified medical documents.

---

🚑 Problem This Project Solves

Healthcare information online is often:

• Scattered across unreliable sources
• Confusing or full of outdated information
• Hard for non-technical users to search
• Lacking verified medical grounding
• Overwhelming due to medical jargon

RxAI solves this by offering:

• A centralized, verified, and search-optimized medical knowledge base
• A simple chat interface where users can ask anything
• AI-generated responses grounded in retrieved medical documents, not hallucinations
• Quick and easy access to medical information
• 24/7 availability with accurate, context-driven responses

---

🌟 Advantages of RxAI

✔ RAG-based accuracy – No hallucinations, responses come from real medical data
✔ Fast & scalable vector search using Pinecone
✔ High-quality embeddings (Sentence-Transformers)
✔ Lightweight architecture using Flask
✔ Easy to extend (add more PDFs anytime)
✔ Secure via environment variables
✔ Clean, modular codebase suitable for production

---

🚀 Getting Started

1 Clone the Repository

  git clone https://github.com/your-username/RxAI.git
  cd RxAI

2 Create a Virtual Environment

  python3 -m venv venv
  source venv/bin/activate      # macOS / Linux
  venv\Scripts\activate         # Windows

3 Install Dependencies

  pip install -r requirements.txt

If you're using Pinecone with gRPC:

  pip install "pinecone-client[grpc]"

4 Set Up Environment Variables

Create a .env file in the project root:

  PINECONE_API_KEY=your_pinecone_api_key
  OPENAI_API_KEY=your_openai_key  # if using GPT-based models
  INDEX_NAME=rxai-index

5 Prepare & Store Your Medical Data

Place all your medical PDFs inside:

  Data/

Then run:

  python store_index.py

This will:

• Load PDFs
• Split documents
• Generate embeddings
• Create (if needed) and upload vectors to your Pinecone index

6 Run the Backend (Flask App)

  python app.py

You should see:

  Flask server running on http://127.0.0.1:5000

🧪 API Usage (Example)

POST /ask

  {
    "query": "What are the symptoms of dengue?"
  }

Response:

  {
    "answer": "Common symptoms of dengue include high fever, severe headache, joint pain..."
  }

🗂️ Project Structure

    RxAI/
    │── app.py                 # Flask backend
    │── store_index.py         # Creates Pinecone index & stores vectors
    │── requirements.txt
    │── .env
    │── Data/                  # Your medical PDFs
    │── src/
    │     ├── helper.py        # Embeddings, loaders, and utilities
    │── README.md

🛠️ Tech Stack

• Python
• Flask
• LangChain
• Pinecone (Serverless)
• HuggingFace Embeddings
• Sentence Transformers
• Any LLM Backend

📌 Future Improvements

• UI frontend (React or Next.js)
• Multi-language support
• More medical datasets
• Voice input + voice output
• Patient history tracking
• Drug-to-drug interaction checker

❤️ Contributing

• Pull requests are welcome!
• Feel free to open issues for bugs or feature requests.

📜 License

MIT License – Free to use, modify, and distribute.