// real interviewer

A face, not a form — and a panel, not one person.

Typing answers into a box teaches you to type answers into a box. Here someone asks you a question out loud, waits, hears what you said, and reacts to it — including following up when your answer was thin. That is the part that transfers to the real room.

4interviewers, each running their own round 11languages, spoken natively 1follow-up per question, like a real probe

The questions are built for this job, not pulled from a list

Every set is written from four things at once: the posting you pasted, your CV, the occupation's real tasks and tools, and live research on the company. A question about "handling scale" for a warehouse supervisor is not the same question as for a backend engineer, and neither is a stock question.

The set also carries what a strong answer would contain — the signals to listen for — which is what the scoring is later measured against.

The interview screen mid-question — The interview screen, mid-question. The question is spoken aloud and printed below the tile so nothing is missed — and there is **no self-view**, because we never turn your camera on.

Four interviewers, because that's how hiring works

A real process is not one person. Yours is a panel, and each round is run by the person who would really run it:

Emily Carter, recruiter — the phone screen. Fast, friendly, checking basics.
Dana Cohen, hiring manager — the general interview. Warm but sharp, cross-checks your CV against what you just said.
Maya Levin, HR business partner — the HR round. Conflict, values, gaps and short stints, asked with zero judgment.
Alex Brandt, staff engineer — the technical round and the written exercise. Brief, precise, pushes on trade-offs you skipped.

Each has their own voice and their own face, and they behave differently on purpose. The HR partner will put you at ease when you sound tense; the engineer will not.

They follow up — which is where interviews are actually lost

Say something vague and you get asked about it, once, right then: "you said it handled a million rows — where did it break first?". One probe per question, then the interview moves on, exactly like a real one. Nobody is trying to trap you; they're trying to find the specifics that a good answer would have contained.

On camera, or just a voice

Choose per position. Video call films the interviewer saying every question — the closest thing to the real screen. Voice + portrait speaks the question over a still photo, which costs a fraction of the tokens and asks exactly the same questions. Phone is a voice-only call for the recruiter screen.

You can also render every video for a position up front, so no round ever pauses to film. And there is no camera on your side — we never ask for one.

Start practicing free See the whole path