A coding interview that pushes back.
Alex gives you a problem and listens while you think out loud. Your code runs against hidden tests you can't see. Pass too easily and the problem gets harder. At the end you get a written scorecard built from what you actually did, not generic praise.
free · no account · 30-45 min · mic recommended
How a round goes
- 1Clarify and plan. The editor stays locked until you've stated an approach and its complexity, same as a real on-site.
- 2Code in a real editor. Run the visible examples yourself; Submit grades hidden edge cases. Brute force that passes gets you one question: can you do better?
- 3Solve it and the problem escalates into a harder variant with new constraints and fresh tests, on the same clock.
- 4Get a scorecard grounded in the evidence: your transcript, how your code evolved, every test run, your pacing, and whether you went quiet while coding.
Why not just ChatGPT?
Chat grades the conversation. This grades the work. It watches the editor, executes your code against tests it never shows you, and hears the difference between narrating your approach and typing in silence. When the feedback says you failed a hidden edge case twice, it is because it was there.