The Design Round Itself
Forty-five minutes, an open-ended prompt, and an interviewer waiting for you to start. The candidates who do well are rarely the ones who know more. They are the ones running a sequence.
Prerequisites: None. This is the first thing to read.
After this: Run a design round on a repeatable sequence instead of improvising, and know where the time goes.
Suggested first pass: Read sections 1–5, answer each section in your own words, then use the remaining failure modes and exercises as the advanced pass.
"Design a ride-hailing app" is not a knowledge question. There is no answer key. The interviewer fills in a scorecard while you talk, and the rows are roughly the same everywhere.
- Scoping: did you cut the problem to something buildable in 45 minutes?
- Justification: does each choice trace back to a requirement, or to taste?
- Depth: did you find the hard part and go into it?
- Buildability: could a team start from what you drew?
Depth is the row that separates levels. A mid-level candidate produces a diagram that is correct. A senior candidate produces the same diagram plus "this breaks when the hot partition takes 40% of writes, and here is what I would do about it." Same boxes, different score.
Candidates rarely lose this round by not knowing what Kafka is. They lose it by spending twenty minutes on the wrong thing: an elaborate high-level diagram, then reaching the interesting problem with four minutes left. The sequence below exists to protect the last fifteen minutes.