Practise coding interviews out loud, with an AI that talks back.

What actually happens

1. The interviewer opensIt introduces itself, asks your name, and answers anything you want to know about the format before you begin.
2. You get a problemIt writes the statement into your editor and talks you through what you are being asked for. Which problem you get is chosen at the start of the session, not by you.
3. You think out loud and write codeSay what you are considering and why, the way you would to a person. The interviewer reads your code as you work, stays quiet while you are making progress, and answers whenever you ask it something. There is no run button — you reason about your code and talk through your test cases, as in a real screen.
4. It pushes on your solutionComplexity, edge cases, the assumption you made without saying it, what changes if the input gets bigger. It will not hand you the answer.
5. It ends the sessionThe interviewer decides when you are done, and you can save your code and the full transcript to keep.

Before you click start

• Set aside about 45 minutesThere is no timer and the interviewer will not rush you, but it is a full interview rather than a quick demo. Start it when you have a clear stretch of time.
• Headphones are recommendedNot just for comfort. On laptop speakers the interviewer hears its own voice through your microphone and cuts itself off mid-sentence. Worth sorting out now rather than after you have started.
• Somewhere you can speak freelyThe whole session is spoken. A room where you would take a phone call is the right setting.
• It is a betaSignalLoop is early and things may not go smoothly. There is a box at the end of the session to tell us what went wrong.
• The session is recordedYour microphone audio, both transcripts and everything you type in the editor are recorded and stored, so please keep sensitive or personal information out of the conversation. We will say this again and ask you to agree before anything is switched on. Read the privacy policy