Interview practice, out loud
Rehearse the interview before it counts. Talk to an AI interviewer that asks the follow-ups you’re dreading, scores your answers against the real rubric, and shows you exactly where you rambled — while there’s still time to fix it.
No account needed for the first question. Voice, not typing.
The problem
Practising interviews means voluntarily being bad at something you’re about to be judged on. Avoidance wins every time — so the first rehearsal ends up being the real interview.
You can’t read your way into composure. Fluency comes from saying the answer out loud, badly, several times — not from a list of things to remember.
Not to what you know. That’s why the third time you tell a story it lands, and the first time it meanders. Reps are the only thing that moves that line.
How it works
Paste the posting and your CV. It builds an interview around the actual role you’re going for — not a generic list off the internet.
It asks, listens and follows up. Vague answers get pushed on. You can interrupt it mid-question. It won’t be polite about a waffled answer.
Rubric-anchored feedback, the full transcript, and the two or three things to fix before the real thing. Read it, then run it again.
The Daily
One question every morning. Ninety seconds. It isn’t a tip and it isn’t a list — it’s a question you actually answer out loud, then get scored on.
The smallest possible way to stop avoiding it.
One email a day. Unsubscribe in a click. No employer ever sees it.
Not another chatbot
You read. You nod. Nothing is rehearsed, and nothing sticks.
You type. And typing is not the skill being tested — speaking under pressure is.
You talk, it pushes back, and you get a score you can act on. Then you do it again tomorrow.
Pricing
Answer one question out loud and see your score. No account.
One full rehearsal, end to end.
Three sessions. One a day until your interview.
Rehearse the interview you actually have.
Coaches, bootcamps and university career services: run your people through the same engine, per seat, with their progress in one place. Talk to us about seats →
Questions
No. ChatGPT has a conversation; this has a rubric. The same questions and the same scoring anchors are applied to everyone practising for that role, so a score means something and the second attempt is comparable to the first. The conversation is the easy part — the judgement is the product.
No. It won’t accept “I’m a people person” as an answer, and it will follow up until you’re specific. That is the entire point: a friendly chatbot does not make you better at being asked something you didn’t prepare for.
The structure is generic — competency questions, follow-ups, pushback — and the questions come from your posting and CV, so it rehearses your actual interview rather than a stock list. Common roles also have reviewed kits, written and checked by people who do that job.
Voice. You speak, it listens, and you can interrupt it mid-question. A face is available on the tailored tier — but the skill being built is what comes out of your mouth under pressure, not how you look doing it.
Your recordings and transcripts are yours. Nothing is shared with an employer, and nothing is used to train anyone’s model without asking. Delete a session and it is gone.
No, and you should distrust anyone who does. We can promise you’ll have said the hard answers out loud several times before they’re asked for real. What you do with that is yours.
Start today and you’ll have answered the hard questions out loud before anyone asks them for real.
Free, no account, ninety seconds.