Built and shipped production healthcare software handling real-time signals, cloud infrastructure, and secure APIs at scale.

• Built and shipped a production Progressive Web App (PWA) enabling contactless health screening using device camera input in under 30 seconds
• Designed and implemented secure, low-latency REST APIs with strict input validation, failure isolation, and monitoring
• Containerized backend services using Docker and deployed on Microsoft Azure with CI/CD pipelines for safe rollouts and rollbacks
• Collaborated with ML and product teams to productionize computer-vision pipelines with reliability, observability, and SLA guarantees
• Integrated backend inference services aggregating 25+ health markers across distributed components

Built and optimized machine-learning pipelines for time-series anomaly detection in a research environment.

• Built an LSTM-Autoencoder pipeline for time-series anomaly detection, improving F1 score from 0.78 to 0.90
• Applied feature engineering and noise filtering to reduce false positives and improve detection reliability
• Optimized inference and preprocessing pipelines, reducing latency by 20% and increasing throughput by 40%
• Analyzed anomalous patterns and refined detection thresholds for behavior-based fault and threat detection
• Produced research-grade experimental results and technical documentation