Optimize Application Performance and Scalability

Provides expert guidance for diagnosing and optimizing application performance - observability, profiling, load testing, caching architectures, and scalability across frontend, backend, and distributed systems.

Use this skill when diagnosing performance bottlenecks in backend, frontend, or infrastructure, designing load tests or capacity plans, setting up observability and performance monitoring, or optimizing latency, throughput, or resource efficiency. Not a fit for feature development with no performance goals, situations without access to metrics/traces/profiling data, or when only a quick non-technical summary is needed. Follows a four-step process: confirm performance goals and baseline metrics, collect traces/profiles/load tests to isolate bottlenecks, propose optimizations with expected impact and tradeoffs, and verify results with regression guardrails - avoiding load testing production without approvals and using staged rollouts with rollback plans for high-risk changes.

Covers observability and monitoring: OpenTelemetry for distributed tracing, APM platforms (DataDog, New Relic, Dynatrace, Honeycomb, Jaeger), metrics stacks (Prometheus, Grafana, InfluxDB, SLI/SLO tracking), Real User Monitoring (Core Web Vitals, page load analytics), synthetic monitoring, and log correlation. Application profiling covers CPU (flame graphs, hotspot identification), memory (heap analysis, GC tuning, leak detection), I/O (disk, network, database query profiling), language-specific profiling (JVM, Python, Node.js, Go), and cloud profiling (AWS X-Ray, Azure Application Insights, GCP Cloud Profiler).

Load testing guidance covers k6, JMeter, Gatling, Locust, Artillery, browser testing (Puppeteer, Playwright, Selenium), chaos engineering (Chaos Monkey, Gremlin), and CI/CD-integrated performance budgets and regression detection. Caching guidance spans application caching, distributed caching (Redis, Memcached, Hazelcast), database caching (query result caching, connection pooling), CDN optimization (CloudFlare, CloudFront, Azure CDN), browser caching (HTTP cache headers, service workers), and API response caching/invalidation.

Frontend optimization covers Core Web Vitals (LCP, FID, CLS), resource optimization (image/lazy loading), JavaScript bundle splitting/tree shaking, critical CSS, HTTP/2/3, and Progressive Web App caching strategies. Backend optimization covers API response time, microservices patterns (circuit breakers, bulkheads), async processing/message queues, database query/index/connection-pool optimization, and concurrency tuning. Distributed system performance covers service mesh tuning (Istio, Linkerd), message queue optimization (Kafka, RabbitMQ, SQS), API gateway rate limiting, and gRPC/REST/GraphQL cross-service optimization.

Cloud performance covers auto-scaling (HPA/VPA), serverless cold-start optimization, container/Kubernetes resource limits, and cost-performance right-sizing. Database performance covers execution plan analysis, connection pooling, NoSQL tuning (MongoDB, DynamoDB, Redis), and time-series storage optimization (InfluxDB, TimescaleDB). Mobile/edge performance covers React Native/Flutter optimization, edge computing, and battery/perceived-performance tuning. Performance analytics covers session replay/heatmaps, performance-revenue correlation, and anomaly-based alerting.

Spans OpenTelemetry, major APM platforms (DataDog, New Relic, Dynatrace, Jaeger), Prometheus/Grafana, load testing tools (k6, JMeter, Gatling, Locust), caching layers (Redis, Memcached, CDNs), and cloud profiling tools across AWS/Azure/GCP.

Performance engineers diagnosing bottlenecks or designing scalable, observable systems who need concrete tooling and architectural guidance across the full stack rather than a single-layer optimization checklist.