Self-Improving Latency: Tail Latency Focus (p95/p99)
Design an agent that optimizes tail latency: identify bottlenecks, reduce contention, add caching, and validate with load tests. Track p95/p99 improvements.
Author: Assistant
Category: safe-self-improving-ai | Model: gpt-5.2