What Is a System Design Interview?

Full Definition

System design interviews became standard for senior+ engineering candidates around 2015 and are now mandatory at virtually every tier-1 tech company for L4/L5 and above. The structure is consistent: the interviewer presents a prompt ('design Twitter', 'design a parking lot', 'design WhatsApp'), the candidate clarifies requirements, estimates scale, sketches a high-level architecture, deep-dives into one or two components, and discusses trade-offs and follow-ups. The signal interviewers look for is not 'correct design' (there isn't one) but: requirement clarification before drawing, quantitative scale estimation, ability to discuss multiple architectural options with explicit trade-offs, awareness of real production systems at similar scale, and the ability to navigate follow-up probes without getting stuck. Common rubric items include: did the candidate ask about read/write ratio? did they discuss consistency vs availability? did they identify the bottleneck component? did they handle a follow-up like 'what if we 10x the scale?' The depth expected scales with the candidate's level: L4 candidates can get away with a clean high-level diagram and one deep-dive; L6+ are expected to discuss operational concerns (deployment topology, monitoring, oncall) on top of the technical design. Preparation typically involves drilling 10-20 canonical problems (TinyURL, Twitter, Uber, Netflix, WhatsApp, rate limiter, distributed cache) until the candidate can draw and defend each in 30 minutes.