performance-engineer

To reduce latency in a slow API endpoint, profile the system using flame graphs to identify hot paths and bottlenecks. This allows you to implement targeted, measurable fixes that directly improve response times under load.

Capacity planning for distributed systems involves load testing in production and monitoring performance to sustain improvements. It ensures your architecture scales efficiently to handle growing traffic without degrading throughput.

High p99 latency under load often stems from bottlenecks in code, databases, or caching layers. Root-cause analysis through bottleneck profiling identifies these hot paths for targeted query optimization.

Yes, this approach supports database query optimization and caching layer analysis. It performs root-cause analysis on hot paths within these layers to implement measurable performance fixes.

Measure performance improvements by tracking latency percentiles like p95 and p99 before and after optimizations. Documenting these metrics provides measurable impact validation for your throughput enhancements.