Led development of a centralized monitoring platform that tracks API performance and system health metrics in real-time; ML-driven anomaly detection reduces incident response time and prevents ~15% of potential downtime events.
Challenge: Amgen's distributed digital infrastructure lacked unified visibility, making it difficult to proactively identify service disruptions that could impact critical pharmaceutical operations and clinical workflows.
Solution: Architected and led development of a full-stack web application featuring interactive dashboards for real-time visualization of API traffic patterns, latency metrics, error rates, and system logs across multiple services. Implemented ML-powered anomaly detection using isolation forests and time-series analysis to automatically surface irregularities and trigger intelligent alerts before issues escalated into incidents.
Impact: Provided 24/7 operational visibility across Amgen's digital ecosystem, reduced mean time to detect anomalies by 30%, and enabled proactive incident prevention through predictive alerting.
Technologies: Java, Spring Boot, Python, Flask, scikit-learn, Pandas, JavaScript, RESTful APIs, real-time dashboards, ML anomaly detection, time-series analysis.