AI Telehealth is full-stack, production-ready AI telehealth software. Built with Python/FastAPI on the backend and React 18/TypeScript on the frontend, it gives you a working codebase you can deploy, extend, and operate as your own closed-source product or SaaS service. It is designed for developers and technical founders who want a substantial head start building a healthcare application without starting from scratch.
The software ships with five integrated AI and ML modules: an NLP-based symptom triage classifier (RandomForest with TF-IDF, four urgency levels), a streaming medical chat assistant (RAG pipeline using FAISS vector search and Flan-T5), an automated SOAP note generator (four-section clinical documentation via Flan-T5), a skin condition image analyzer (MobileNetV3 architecture supporting 15 dermatological classes), and a multi-factor health risk assessor for cardiovascular and metabolic scoring. Beyond the AI layer, the software includes appointment booking and management, consultation records, a digital prescription manager, and separate role-based dashboards for patients and doctors, all secured with JWT authentication.
Training data for the triage classifier (400 entries, four urgency classes) and the medical knowledge base (200 entries across 10 specialties) is synthetically generated. Synthetic data is used deliberately to improve class coverage and to ensure no real patient information is embedded in the codebase or distributed with the product.
Requirements: Python 3.11, Node.js 18 LTS, npm 9+, and at least 4 GB of free disk space. The first startup downloads approximately 2 GB of AI model weights; a stable internet connection is required. SQLite is included for local development; PostgreSQL with asyncpg is recommended for any production deployment. The skin condition analyzer ships in demo mode by default and returns simulated results until you supply a custom-trained MobileNetV3 model file (training instructions and the expected model architecture are included in SETUP.md and TRAINING_DATA.md).
