reliability
Distributed systems that fail safely
Retries, idempotency, failover, timeouts and the observability to prove they work. The playground above is the short version of what I do for a living.
- Wrote an idempotency starter for Spring Boot connectors: contract-driven, Redis-backed, about 96 tests
- Root-caused production OOM kills and tuned JVM memory, probes and autoscaling on Kubernetes