The system
VitaLink AI is a real-time health-monitoring project developed from October 2025 to January 2026. It connects ESP32 sensor data to responsive web interfaces for live monitoring, historical review, data visualization, and administration.
The implementation
I developed a landing page, student dashboard, and admin dashboard with Next.js and Tailwind CSS. A FastAPI backend handles authentication and communicates with the ESP32 and frontend through WebSockets, providing secure access and real-time sensor streaming.
Anomaly detection
An Isolation Forest model built with Scikit-Learn processes sensor data to identify abnormal heart-rate and motion patterns. Detected anomalies surface as health alerts in the dashboards alongside live and historical readings.