Skip to content

🏗️ System Design Interview

End-to-end design practice for Staff-level rounds. A reusable framework, an estimation cheat sheet, and eight case studies that each follow the same structure so you can compare how decisions change with the problem.

Start here

Step Read Why
1 Framework & Estimation The 45-minute structure, numbers to memorize, and what "Staff" signals look like
2 Any case study below Practice the framework on a concrete problem
3 CS Core: Distributed Systems The theory the deep dives lean on (consensus, CAP, consistent hashing)
4 Staff Craft: Reliability & Operations The "failure modes and rollout" half of every design

Case studies

# Case study The hard part
1 URL Shortener Collision-free code allocation, redirect hot path, click counting
2 News Feed Hybrid fan-out (celebrities), stable pagination, deletes and unfollows
3 Chat & Messaging 100M concurrent connections, ordering, offline delivery
4 Typeahead / Search Suggestions Sub-100 ms prefix lookup, offline aggregation, freshness
5 Ad Click Aggregation Exactly-once-effective counting, late data, hot ads
6 Video Streaming Transcoding DAG, adaptive bitrate, CDN economics
7 File Storage & Sync Chunking and dedupe, ordered metadata, conflicts
8 Payments Ledger Double-entry, idempotency, unknown outcomes, reconciliation

Every case study has the same 11 sections

Requirements → Estimation → API → Data model → High-level design → Deep dives → Failure modes → Observability & rollout → Alternatives → Follow-up questions → Common mistakes.

How to practise: read only sections 1–2, close the page, and design on paper for 30 minutes. Then compare your deep dives and trade-offs with section 6 and your follow-up answers with section 10.