What this prompt does
This prompt turns the AI into an interactive system design interviewer. You set a [target_role] and a [company_type], and it gives you a design problem at the right level, then runs a 45-minute mock interview that mirrors a real loop. It follows five rules: open with clarifying questions, push back on vague answers, introduce constraints mid-interview, ask about scaling and failure modes around minute 35, and close with specific feedback plus a hire/no-hire signal — without revealing solutions.
The interactivity is the point. A static problem lets you hand-wave your way past weak spots; a back-and-forth that pushes back exposes the gaps you would never catch reading a solution alone. Tuning the difficulty to [target_role] and [company_type] keeps the problem realistic — a Staff Backend prompt at a FAANG-tier company probes deeper into scale and tradeoffs than an early-career one would. Adding constraints mid-stream forces you to adapt a design under pressure, which is exactly what a real interviewer does, and withholding the solution means you cannot lean on a model answer when you stall.
When to use it
- Preparing for a system design loop at a specific
[company_type] - Practicing the clarify-first habit instead of jumping straight to architecture
- Pressure-testing whether you can defend tradeoffs when challenged
- Rehearsing the scaling-and-failure-modes discussion that closes most loops
- Getting an honest hire/no-hire signal before the real interview
- Calibrating to a
[target_role]level you are stretching toward
Example output
The session unfolds as a real conversation: it states a problem, asks you clarifying questions, reacts to your answers, and injects a constraint partway through. Near minute 35 it pivots to scaling and failure modes. At the end you get specific feedback on what was strong and weak, plus a hire/no-hire call. It deliberately withholds the model solution so you cannot lean on it.
Pro tips
- Set
[target_role]to the exact level you are interviewing for so the problem difficulty matches — a Staff prompt assumes more than a Senior one - Use
[company_type]to calibrate expectations; "FAANG-tier" pushes harder on scale than a small-startup framing - Actually answer out loud or in full text — terse replies invite the push-back rule, which is where the value is
- When it introduces a mid-interview constraint, adapt your design openly rather than defending the original; that is what it scores
- Ask it to be harsher on the scaling section if your weak spot is failure modes
- Run it twice with the same
[target_role]and compare the two hire/no-hire signals to spot consistent gaps - Treat the clarifying-questions phase as scored too — jumping straight to a design without scoping requirements is a common reason real loops downgrade a candidate
