The most structured round. Driving the conversation top-to-bottom IS the grade.
~42 min · run it top to bottom
- 1. Functional requirements5 min5:00
Clarify scope. Pin the 1–3 core use cases out loud and explicitly cut the rest for now.
Interviewer grades: Do you scope it, or try to boil the ocean?
Say it in your own words — but hit these: core use case · out of scope · the main flow
checking microphone…
- 2. Non-functional + scale5 min5:00
Scale numbers, latency target, consistency vs availability. A back-of-envelope estimate.
Interviewer grades: Do you quantify, and do you name the consistency model?
- 3. Core entities + API7 min7:00
The data model and the key endpoints (resource URLs, the transition endpoint).
Interviewer grades: A clean data model and a resource-based API.
- 4. High-level design8 min8:00
Client → API → service → Postgres + cache + queue + workers. Draw it on the canvas.
Interviewer grades: One coherent end-to-end path, components earning their place.
- 5. Deep dive12 min12:00
The 1–2 hard parts: the bottleneck, the failure modes, how it scales 10×.
Interviewer grades: Depth on the hard part — this is where the round is won or lost.
- 6. Tradeoffs + monitoring5 min5:00
Name the tradeoff for each choice and what you'd monitor/alert on.
Interviewer grades: Tradeoff reasoning and operational maturity.
Say it in your own words — but hit these: source of truth · idempotent · outbox · async workers
checking microphone…