Voice mock interviews for AI and ML engineers.

What a voice mock is

A voice mock interview is a timed AI/ML technical mock you answer out loud instead of typing. The examiner listens to your answer, asks a follow-up as soon as you finish speaking, and grades the transcript against the same fixed criteria as a text mock. Live captions show what the examiner heard, and you can switch to text at any point in the sitting.

How a spoken sitting runs

1. Choose your role, seniority and two to four questions; the time budget is shown before you start.
2. The microphone stays off until you start it.
3. You answer each question aloud.
4. The follow-up arrives as soon as you finish speaking.
5. Live captions show what the examiner heard.
6. You can switch to text at any point in the sitting.
7. The timer runs with no hints and no pauses.
8. After the last question, you get the debrief.

What is graded, and what is not

The examiner grades the reasoning in your transcript against fixed criteria that are reviewed before the mock and cannot change mid-interview. Feedback cites evidence from your recorded answers. When the evidence is too thin to support a finding, the report marks it inconclusive rather than guessing, and topics the sitting did not reach stay marked untested.

Because grading works from the transcript, delivery is not part of the score: accent, pace and filler words are not graded.

An illustrative spoken exchange

Illustrative example. This exchange is the same teaching example used elsewhere on this site to show how a follow-up works. It is not an actual learner transcript or a generated product report. Actual mocks vary.

EXAMINER

Your classifier is 99% accurate. Is it ready to ship?

YOU

The accuracy is high, so I would deploy it.

EXAMINER · FOLLOW-UP

Only 1% of examples are positive. What would always predicting negative score, and what would you measure next?

Voice or text: how to choose

Both modes run the same mock and are graded the same way, on the reasoning in your transcript. In text, you write your answer, with support for equations and a notebook. In voice, you speak, the examiner speaks back, and captions show what it heard as you go. The free first mock always runs in text. Choose voice if your actual interviews are spoken and you want to practise thinking out loud under a clock.

Plans

Voice mocks require the Voice plan; Text covers unlimited written mocks on its own.

First completed mock free, in text, no card required. See current plans and prices.

Keep exploring

See how the same grading approach works for a specific role, or read a full example report end to end.