03-perf-optimization

Example 3: Performance Optimization

Goal: Reduce API response time from 450ms to under 200ms Metric: p99_latency_ms (direction: lower) Verify: node benchmark.js --metric p99 Guard: npm run test:integration

Scenario

A Node.js API endpoint has slow response times (450ms P99). AutoResearch systematically tests optimization approaches—caching, database indexing, query optimization—measuring each with real-world load testing.

Expected Outcome

• Iterations: 12
• Best latency: 187ms
• Kept: 4, Discarded: 8

Run Initialization

autoresearch init \
  --goal "Reduce /api/users endpoint P99 latency from 450ms to under 200ms" \
  --metric "p99_latency_ms" \
  --direction "lower" \
  --verify "node benchmark.js --metric p99" \
  --guard "npm run test:integration" \
  --iterations 20 \
  --max_no_progress 6

Files

• state.json — Run checkpoint at completion
• results.tsv — Iteration log with decisions
• report.md — End-of-run summary