title	NewGenHealthAI
emoji	🩺
colorFrom	blue
colorTo	green
sdk	docker
pinned	false
app_port	7860

🩺 NewGenHealthAI

An AI-powered Medical Consultation & Diagnostic Platform

🚀 Live Demo

https://huggingface.co/spaces/sg721642/NewGenHealthAI

---

An AI-powered medical assistant with a FastAPI backend and a React frontend for conversational clinical guidance, session-based chat, and image-based disease analysis. Deployed and publicly accessible through Hugging Face Spaces.

Overview

NewGenHealthAI is a full-stack healthcare AI application designed to support medical interaction workflows through:

• conversational AI assistance
• image-based disease analysis
• session-based chat history
• diagnostics and model inspection utilities
• containerized deployment with Docker

The project is organized as a monorepo with separate backend and frontend applications.

Repository Structure

NewGenHealthAI/
├── backend/
│   ├── app/
│   ├── test_images/
│   ├── tests/
│   ├── .flake8
│   ├── Dockerfile
│   ├── inspect_models.py
│   ├── pyproject.toml
│   ├── pytest.ini
│   ├── requirements.txt
│   ├── test_api.py
│   └── test_clip_routing.py
├── frontend/
│   ├── public/
│   ├── src/
│   ├── .gitignore
│   ├── Dockerfile
│   ├── README.md
│   ├── eslint.config.js
│   ├── index.html
│   ├── nginx.conf
│   ├── package-lock.json
│   ├── package.json
│   ├── postcss.config.js
│   ├── tailwind.config.js
│   └── vite.config.js
├── .env.example
├── Dockerfile
├── check_model_on_image.py
├── diagnose_env.py
├── run.py
├── run_medai.bat
└── .gitignore

Features

• AI-powered medical assistant interface
• FastAPI backend API
• React + Vite frontend
• Session-based conversation history
• Medical image analysis endpoint
• Model diagnostics endpoint
• Local testing utilities for medical images
• Docker support for deployment
• Environment-based configuration

Tech Stack

Backend

• Python
• FastAPI
• Uvicorn
• LangChain / LangGraph
• ChromaDB
• Sentence Transformers
• Hugging Face Transformers
• PyTorch
• SQLAlchemy

Frontend

• React
• Vite
• Tailwind CSS
• DaisyUI

Environment Variables

Create a .env file in the project root based on .env.example.

Required variables include:

GROQ_API_KEY=your_groq_api_key_here
TAVILY_API_KEY=your_tavily_api_key_here
SESSION_SECRET_KEY=your_session_secret_key_here

Optional path overrides may also be configured for logs, database, vector store, and PDF sources.

Installation

1. Clone the repository

git clone https://github.com/sg721642/NewGenHealthAI.git
cd NewGenHealthAI

2. Backend setup

cd backend
python -m venv .venv

Activate the virtual environment:

Windows

.venv\Scripts\activate

Linux / macOS

source .venv/bin/activate

Install dependencies:

pip install -r requirements.txt

3. Frontend setup

Open a new terminal:

cd frontend
npm install

Running the Project

Option 1: Start with the root launcher

From the project root folder:

python run.py

This script:

• creates a virtual environment if needed
• installs backend dependencies
• installs frontend dependencies
• starts the backend on port 8000
• waits for the backend health endpoint
• launches the frontend development server

Option 2: Run backend and frontend separately

Start backend

cd backend
uvicorn app.main:app --reload --port 8000

Start frontend

cd frontend
npm run dev

API Endpoints

Some important backend routes include:

• GET /api/v1/health — backend health check
• GET /api/v1/diagnostics — loaded model diagnostics
• POST /api/v1/analyze-image — upload an image for analysis

Depending on your backend router definitions, additional chat and session endpoints are also available.

Testing

Backend tests

Run pytest inside the backend folder:

cd backend
pytest

Image API test script

cd backend
python test_api.py

CLIP routing verification

cd backend
python test_clip_routing.py

These scripts test image classification and verify that domain routing works correctly for sample images such as skin, xray, and retina inputs.

Docker

Dockerfiles are included for:

• project root
• backend
• frontend

You can use them to build and containerize the entire application or individual services depending on your deployment setup.

Frontend

The frontend is built with React and Vite and provides:

• chat-based interaction
• conversation history
• a clean medical assistant interface

To create a production build:

cd frontend
npm run build

To preview the production build locally:

npm run preview

Utilities

The project also includes utility scripts such as:

• check_model_on_image.py — test model behavior on a single image
• diagnose_env.py — inspect environment setup and issues
• inspect_models.py — inspect backend model loading/configuration
• run_medai.bat — Windows launcher helper

Use Cases

• AI-assisted symptom guidance
• educational medical assistant demos
• image-based disease analysis workflows
• research prototypes for healthcare AI systems

Disclaimer

This project is intended for educational, research, and prototype purposes only.

It must not be used as a replacement for licensed medical diagnosis, treatment, or professional clinical judgment.

Contributors

• Satyam Gupta
• Khagesh Ranjan
• Sudipto Ghosh
• Ajitesh Baghel
• Yuvraj Singh

License

Add your preferred license here, for example:

MIT License

If you have not chosen one yet, create a LICENSE file before publishing for open-source reuse.

Acknowledgments

Special thanks to the mentor, contributors, and open-source communities behind FastAPI, React, Vite, Hugging Face, LangChain, and related AI tooling.