Backend engineer specializing in performance engineering and JVM internals.
Diagnoses concurrency bottlenecks through distributed metrics (Grafana) and low-level profiling (JFR/Thread Dumps) with hypothesis-driven analysis.
SLAM! — Concert Seat Reservation System handling high-concurrency bookings with real-time availability and payment processing
- Performance: P99 latency 1.04s → 355ms (-66%), throughput +54% — root-caused G1GC STW pauses via JFR analysis and Little's Law verification
- Reliability: 3-layer API idempotency (Redis Response Cache → Redisson Distributed Lock → DB Unique Constraint)
- Capacity Control: Redis Sorted Set waiting queue with 90% network RTT reduction via pipelining
- Observability: Prometheus, Grafana, Loki with Trace ID correlation and threshold alerting
- Languages & Frameworks: Java 17, Spring Boot 3, Spring Data JPA, JUnit 5, Python
- Data & Infrastructure: MySQL 8.0, Redis 7.0, AWS (EC2, RDS, ECR, S3), Docker, GitHub Actions
- Observability & Analysis: Prometheus, Grafana, Loki, JFR, Thread Dump Analysis