You cannot rehearse an interview alone. This one asks out loud.

Which format is yours

Two modes of one product. The same mocks pay for both, and every scorecard names the other mode as the next step, because most people face both formats in the same search, a few weeks apart.

Set harder than the real interview, on purpose

The interviewer does not take the first version of an answer. Say you improved something and it wants the number. Say the team shipped it and it asks what you did. When an answer is good it presses once more anyway, for what the work cost you or what you would do differently. A real panel lets a good answer go; this does not.

Hard is the default on every run, free ones included. You rehearse above the bar, so the room you walk into is the easier version.

What the simulator scores, in both formats

A form rejection tells you nothing and a friend saying it was good tells you less. Every answer here gets the same five labels, which is what makes a second run comparable to the first. Each row carries the label, a score, and what the model heard that earned it.

• Answer Structure. Interviewers listen for Situation, Task, Action, Result. An answer with no concrete outcome reads as an unproven claim. The Result is what separates a story from an assertion.
• Judgment & Problem-Solving. How you reason through a problem: what you check first, when you escalate or ask for help, and why. Sound reasoning under uncertainty is what a scenario question is actually testing.
• Specificity & Evidence. Concrete details prove the experience is real: the project, the numbers, what YOU personally did. Generic "I would" answers score low because anyone can say them.
• Role Relevance. Whether your answers speak to this job, not a job in general. Interviewers are listening for someone who understood what the role actually involves.
• Communication & Presence. Clarity, pacing, and composure. Rambling, trailing off, or losing the thread costs you even when the content underneath is right.

Every answer also gets a transcript-measured delivery line: seconds spoken, words per minute, filler count. That part is arithmetic on the transcript, not a model's opinion, and the scorecard labels it that way.

What a simulation is not

• Not your employer's questions. Nobody outside the employer has those. The conversational interviewer builds its questions from the role you name or the posting you paste; the recorded interview uses question sets built from what employers publish, labeled with their source.
• Not a replica of any vendor's software. The recorded interview simulates the format the vendors share, a question and a one-take timer, with settings you choose, because the real settings are chosen by your employer and published to nobody.
• Not a person. The interviewer is software with a drawn face, labeled AI on every screen. The scorer is a model reading a transcript. The delivery numbers are arithmetic.
• Not a way to change a hiring decision. It is rehearsal. Rehearsal changes what you say, and that is all it changes.

What is free and what is paid

• Free: Two scored mock interviews, each with your overall score, all five skill scores and the full transcript. No card. When your first scorecard is ready we ask for an email, so you can find it again. Plus two three-minute drills.
• Monthly plan, $19 a month: As many scored mocks as you want, every month. Plus the coaching: why you lost points on each skill, your weakest answers rewritten with the reasons the new version scores higher, and drills for what you missed.
• Renews at $19 a month until you cancel. Cancel from your account page. The next month is not charged. Refunds: 14 days from your first payment, no reason required, one button on your account.

Run it

Another week of searching costs you a week of pay, and the interview is the part of it you control. Name the job, allow the microphone, and the first question comes in under a minute. No account needed to start.