Real-Time Medical Machine Learning Data Pipeline

Design a scalable data pipeline architecture that enables real-time ingestion of medical sensor and diagnostic data for machine learning model training. Create a solution supporting streaming data processing, feature engineering, and automated model retraining with built-in data validation, anomaly detection, and compliance monitoring. Address challenges of handling high-velocity medical telemetry from IoT devices, wearables, and clinical monitoring systems.