AI and your assessment

Your written answers are graded by an AI model against criteria written for this role. Each answer gets a score from 0 to 100 for accuracy, depth and practical understanding.

Your overall score is a plain average of those per-question scores, calculated by our own code rather than by the AI. The thresholds are published: 75 and above is Verified, 50 to 74 is Partially Verified, below 50 is Unverified.

The full grading method, including what we deliberately exclude, is written up in How we score.

• We do not use a camera or microphone at any point. We do not process video or audio, and we do not analyse your face, voice, accent or appearance.
• We do not infer your emotions, personality or mental state.
• We do not score you on writing polish, grammar, or how fluent your English sounds. The grading instructions explicitly exclude presentation, length and formatting.
• We never send your name, email or photo to the model as separate information. The model sees the role, the skills being assessed, and the text of your answers.
• We do not sell your data and we do not use it for advertising. There is no advertising or analytics tracking in this product.
• We do not decide whether you get the job. A person at the company hiring you does.

We record whether text was pasted into an answer, whether it appeared faster than a person can physically type, and how many times the browser tab was switched away from.

These are shown to the recruiter as context for a follow-up question. They never change your score — the grading model is never told about them — and they are not treated as proof of anything on their own.

Tell us and we will pass it to the hiring company, who can offer you a different way to demonstrate your skills. The same route is there if you need an adjustment for a disability.

Request an alternative or an adjustment

Asking does not disadvantage your application.

You can ask for a copy of everything we hold about you — your answers, your per-question scores, the criteria applied, the verdict, and the integrity signals. You can ask for corrections, ask for an explanation, and ask us to delete your assessment.

Requests are normally handled by the company that asked you to take the assessment, since they decide how your results are used. If you cannot reach them, contact us and we will help.

Make a request about your data

Large language models carry subtle preferences, and the same answer can score slightly differently on different runs. We do not claim otherwise.

That is exactly why a person makes the decision, and why we show them your actual answers rather than just a number.