🩺 MediBot – AI-Powered Medical Chatbot (Mobile + Flask Backend)

MediBot is an intelligent, generative AI-based medical chatbot that provides accurate, context-aware responses to health-related queries. It integrates state-of-the-art large language models (LLMs) using Retrieval-Augmented Generation (RAG) and delivers real-time answers via an Android app built with WebView.

⚠️ Disclaimer: MediBot is an informational tool and not a substitute for professional medical advice, diagnosis, or treatment.

---

📌 Features

• 🧠 Powered by transformer-based LLMs (e.g., LLaMA-3.1-8B-Instruct)
• 🔍 Context-aware responses using FAISS + LangChain RetrievalQA
• 📱 Android frontend with React Native + WebView
• 🔗 Seamless Flask-based backend integration
• 📂 Dynamic PDF ingestion and semantic embedding creation
• 🌐 Deployed via Hugging Face Inference API

---

📐 Architecture Overview

flowchart TD
    A[User Query via Android App] --> B[POST to Flask API]
    B --> C[Retrieve Env Vars & Model Config]
    C --> D[LangChain RetrievalQA Pipeline]
    D --> E[Fetch Chunks from FAISS]
    E --> F[Form Prompt + Context]
    F --> G[Call HuggingFace InferenceClient]
    G --> H[Generate & Return Answer]
    H --> I[Display Response in Mobile UI]
    subgraph Setup
    J[Upload PDFs] --> K[Chunk Text] --> L[Embed with SentenceTransformer] --> M[FAISS Index]
    end
    M --> E

Loading

---

🛠️ Tech Stack

Layer	Technology
Frontend	React Native, WebView, Expo Go
Backend	Flask, Python, LangChain, dotenv
Embedding	sentence-transformers, FAISS
LLM	Hugging Face InferenceClient (LLaMA)
Retrieval	LangChain RetrievalQA Chain
Storage	Local FAISS Vectorstore

---

📂 Folder Structure

├── medibot-app/                           # Core logic for embedding + memory management
│   ├── connect_memory_with_llm.py         # Retrieval pipeline with LLM
│   ├── create_memory_for_llm.py           # Embedding medical PDFs into FAISS
│   ├── file.gitignore
│   ├── medibot.py                         # Main interface file
│   ├── test.py
│   ├── README.md
│   ├── requirements.txt
│   ├── data/                              # PDF files for ingestion
│   └── vectorstore/                       # FAISS index storage
│
├── medibot-flask/                         # Flask backend for serving the LLM pipeline
│   ├── app.py                             # Flask API entrypoint
│   ├── requirements.txt                   # Python dependencies
│   ├── .gitignore
│   └── vectorstore/                       # Shared FAISS index
│
├── medibot-react-app/
│   └── medibot-app/                       # React Native frontend
│       ├── App.js                         # Root of the mobile app
│       ├── app/                           # Navigation and screen components
│       ├── assets/                        # Images, icons
│       ├── components/                    # UI components
│       ├── constants/                     # Static configs
│       ├── hooks/                         # React hooks
│       ├── scripts/                       # Any automation/util scripts
│       ├── node_modules/
│       ├── eas.json                       # Expo app config
│       ├── app.json                       # Project manifest
│       ├── package.json                   # JS dependencies
│       ├── tsconfig.json                  # TypeScript config
│       └── README.md

---

🚀 Getting Started

1. Clone the Repository

git clone https://github.com/<your-username>/medibot.git
cd medibot

2. Setup the Backend

cd medibot-flask
python -m venv venv
source venv/bin/activate  # or venv\Scripts\activate on Windows
pip install -r requirements.txt

Create a .env file:

HF_TOKEN=your_huggingface_api_token
HUGGINGFACE_REPO_ID=meta-llama/Llama-3.1-8B-Instruct
DB_FAISS_PATH=vectorstore/db_faiss

Then run:

python app.py

3. Generate the Embedding Memory

python create_memory_for_llm.py

4. Launch the React Native App

cd ../medical-chatbot
npm install
npx expo start

Scan the QR code on Expo Go app to preview it on your phone.

---

📸 Screenshots

Insert screenshots of:

• The Android app UI
• 

• Web App UI
• 

---

🧪 Testing

• Use sample PDFs in the /data/ folder to simulate ingestion.
• Try queries like:
  • “What are the symptoms of dengue?”
  • “How to treat a mild fever?”
• Evaluate response speed, context relevance, and coherence.

---

🔄 Future Enhancements

• 🔁 Multi-turn chat memory
• 🌍 Multilingual input/output support
• 🎤 Speech-to-text and text-to-speech capabilities
• 💾 On-device inference using quantized models
• 📊 Feedback loop + model fine-tuning dashboard

---

🙋‍♀️ Contributing

1. Fork this repo
2. Create a new branch (git checkout -b feature/xyz)
3. Commit your changes (git commit -m 'Add feature')
4. Push to the branch (git push origin feature/xyz)
5. Create a pull request!

---

📜 License

This project is licensed under the MIT License.

---

🙏 Acknowledgements

• Hugging Face
• LangChain
• Sentence Transformers
• Facebook FAISS

Demo Video of the App

• ▶️ Click to watch demo video