Implemented an IoT-based predictive health monitoring pipeline using Python, scikit-learn, and TensorFlow, achieving an AUC of 0.98 and 94% validation accuracy on vital-sign anomaly detection for heart rate, body temperature, and accelerometer data.
Created a scalable real-time data ingestion and remote monitoring platform leveraging MQTT, Flask RESTful APIs, and AWS Lambda functions, enabling continuous data collection, low-latency analytics, and automated alert notifications for abnormal health conditions.
Engineered robust embedded firmware on ESP32 and Arduino microcontrollers to interface with heart-rate, temperature, and accelerometer sensors, orchestrating seamless BLE communication.
