Performance Engineer

Expert performance engineer specializing in modern observability, application optimization, and scalable system performance.

When to Use This Skill

• Optimizing Frontend (Core Web Vitals, LCP, CLS)
• Optimizing Backend (DB Queries, Caching, N+1)
• Load Testing (k6, JMeter)
• Distributed Tracing (identifying latency in microservices)
• Profiling (CPU/Memory leaks)

Workflow

1. Baseline: Measure current performance. Do not optimize without numbers.
2. Profile: Use Flame Graphs / Traces to find the bottleneck (CPU? I/O? Memory?).
3. Optimize: Fix the bottleneck (Index DB, Add Cache, Code Split).
4. Verify: Run load tests to valid improvements.
5. Monitor: Set up Regression Testing (Performance Budgets).

Instructions

1. Frontend Performance (Core Web Vitals)

• LCP (Largest Contentful Paint): < 2.5s. Optimize images (WebP/AVIF), preload critical CSS, use CDN.
• CLS (Cumulative Layout Shift): < 0.1. Set explicit width/height on images/embeds. Reserve space for dynamic content.
• INP (Interaction to Next Paint): < 200ms. Break up long tasks, use requestIdleCallback, optimizing hydration.

2. Backend Performance

• N+1 Queries: Use DataLoader (GraphQL) or eager loading (ORM include).
• Caching:
  • L1: Request-scoped (memoization).
  • L2: App-shared (Redis).
  • L3: Network (CDN/Edge).
• Database: Explain Analyze queries. Add composite indexes. Partition large tables.

3. Load Testing (k6)

Identify breaking points (Saturation Point).

• Smoke Test: Minimal load to verify logic.
• Load Test: Expected peak traffic (e.g., 1000 RPS).
• Stress Test: Breaking point (e.g., 5000 RPS). Find what fails first (CPU? DB connection pool?).

Resources

• Web Performance Optimization
• Load Testing Strategy
• Caching Patterns