Performance Engineering
Where senior engineers separate from the rest. Core Web Vitals, the RAIL model, JavaScript parse cost, network optimization, bundle analysis, and scientific performance methodology.
Think about performance the way Google engineers do. Master the RAIL model, understand the true cost of JavaScript, build performance budgets, and learn why perceived performance matters more than raw numbers.
The three metrics Google uses to rank your site. Master Largest Contentful Paint, Cumulative Layout Shift, and Interaction to Next Paint — what triggers them, how to measure them, and how to fix them in production.
Why smooth animation demands fitting all work into 16.67ms, how the rendering pipeline consumes that budget, and the techniques that let you hit 60fps consistently — compositor-only animations, layout isolation, and scientific measurement.
Why your Lighthouse score lies to you. Understand the fundamental difference between synthetic lab measurements and real-user field data, when to use each, and how to implement RUM with the web-vitals library.
Why JavaScript is the most expensive asset on your page. The full cost breakdown from download to parse to compile to execute — and how to measure, reduce, and eliminate unnecessary main thread work.
JavaScript is single-threaded, but your CPU isn't. Web Workers let you offload heavy computation to background threads — image processing, search indexing, WASM, JSON parsing — without blocking the main thread. Master dedicated Workers, SharedArrayBuffer, Atomics, Transferable objects, Comlink, and when the overhead isn't worth it.
Every millisecond between click and pixels comes from somewhere. Master the critical rendering path, resource hints, HTTP/2 multiplexing, compression algorithms, font strategies, and modern image optimization — the full network performance toolkit that separates senior engineers from everyone else.
From browser memory cache to CDN edge nodes, understand every caching layer that stands between your user and your server. Master Cache-Control, ETags, Service Worker strategies, the Cache API, and Next.js caching — then know exactly when to bust the cache.
Your users download, parse, and execute every byte of JavaScript you ship. Bundle splitting breaks monoliths into parallel-loaded chunks. Tree shaking eliminates dead code at build time. Together they're the difference between a 2s load and a 12s load.
A step-by-step field guide to diagnosing and fixing LCP, CLS, and INP in production. No theory fluff — just the debugging workflows and code fixes that actually move the numbers.
12 scenario-based performance diagnosis questions. Real-world LCP, CLS, and INP problems pulled from production sites. If you can diagnose all 12, you think like a performance engineer.