Frontend System Design
System design content online is almost all backend — load balancers, databases, message queues. The frontend-specific version — rendering strategy, state at scale, performance, micro-frontends — is what a senior or staff frontend interview actually spends its system design round on, and it is rare to find taught well. This is that course.
36 lessons · 448 min total · free to read, no account needed
Rendering strategy
CSR, SSR, SSG, ISR and streaming — the actual trade-offs in TTFB, TTI, SEO and infrastructure cost, not just definitions.
State at scale
Why "put everything in global state" fails, server state vs. client state, and the caching/staleness problem that creates.
Performance at scale
The critical rendering path, code-splitting trade-offs, virtualization, and Core Web Vitals as the actual scoring rubric.
Architecture at scale
Micro-frontends and design systems — what they really cost a team, and when a well-organized monolith is the right call instead.
Practice interview questions
Nine full worked scenarios — the design-X prompts frontend rounds actually ask. Each one walks the reasoning end to end: clarify, propose, trade-offs, scope, takeaways.
- Framework11 min
How to Approach a Frontend System Design Interview
The repeatable approach every answer below uses.
Read the walkthrough - Classic FE14 min
Worked Scenario: Design a Social Media Feed
Infinite scroll, normalized feed data, optimistic likes with a real rollback path.
Read the walkthrough - Classic FE13 min
Worked Scenario: Design an Image-Heavy E-Commerce PLP
ISR for SEO-sensitive listings, image strategy against LCP/CLS, URL-driven filter state.
Read the walkthrough - Classic FE14 min
Worked Scenario: Design a Real-Time Collaborative Editor's Frontend
Why the editor is CSR, optimism at typing-latency stakes, CRDTs for concurrency.
Read the walkthrough - Universal14 min
Worked Scenario: Design a Search Autocomplete (Google-shape)
Debounce + cancellation + match-on-response, bounded LRU, combobox ARIA.
Read the walkthrough - Netflix / YouTube15 min
Worked Scenario: Design a Video Player (Netflix / YouTube)
Adaptive bitrate (HLS/DASH), DRM via EME, custom controls, WebVTT captions, QoE.
Read the walkthrough - Datadog / Grafana15 min
Worked Scenario: Design an Analytics Dashboard (Datadog / Amplitude)
Progressive query loading, cancellation on range change, SVG/canvas/WebGL by density.
Read the walkthrough - Figma / Miro16 min
Worked Scenario: Design an Infinite Canvas (Figma / Miro / Excalidraw)
Spatial index + viewport culling, one-matrix pan+zoom, local-first drag with throttled sync.
Read the walkthrough - Atlassian15 min
Worked Scenario: Design a Kanban Board (Jira / Trello)
Per-column virtualization, concurrent drag reconcile, filter state in the URL.
Read the walkthrough - Atlassian15 min
Worked Scenario: Design a Collaborative Wiki (Confluence)
Progressive tree loading, CRDT/OT for live editing, autosave with real merge.
Read the walkthrough - GitLab15 min
Worked Scenario: Design a Code Review Diff Viewer (GitLab / GitHub)
Two-level virtualization, highlighting in a Worker, comments anchored across force-pushes.
Read the walkthrough - GitLab15 min
Worked Scenario: Design a CI/CD Pipeline Dashboard (GitLab / GitHub Actions)
Layered-DAG stage layout, bounded live-tail log, server-filtered run history.
Read the walkthrough - Microsoft16 min
Worked Scenario: Design a Browser Code Editor (Monaco / VS Code Web)
The editor owns its render loop, tokenization split between main thread and Worker, sandboxed extensions.
Read the walkthrough - Microsoft15 min
Worked Scenario: Design a Chat With Threads (Teams / Slack)
Variable-height virtualization with pinned bottom, threads as a parallel URL, honest presence.
Read the walkthrough
Taught as reasoning, not a checklist
Real system design has few universally correct answers — the skill being tested is whether you can name the trade-offs and argue for one under the constraints you were given. Every lesson here is built around that: what changes if traffic is huge, if the team is small, if SEO matters, if the data changes every second. The capstone lessons pose an actual “design X” prompt and walk through the reasoning a strong answer would use.
- Rendering, state, performance and architecture, covered as trade-offs with real conditions attached
- Diagrams for every architecture and data-flow, not walls of prose
- Capstone lessons that pose a real interview prompt and reason through it end to end
Prove it: Frontend System Design Certification
165 min, 65 questions, a certificate with a public link if you pass.
See the exam