Most hiring tools promise fairness and then fall back on keyword counting. A CV that happens to use the right nouns gets through, and a stronger candidate who described the same work in plain language gets filtered out. XpressApply starts from a different place. Every interview is scored against a written rubric that the hiring team agreed on before anyone applied.

The rubric comes first

Before a role goes live, the company defines the criteria that actually predict success, things like correctness under failure, data modelling, or clear technical communication. Each criterion has a description of what a strong answer sounds like. During your interview the AI is not guessing whether it likes you. It is checking your answers against those descriptions and noting the evidence.

That evidence is the important part. When you get a scorecard, every criterion links back to something you actually said. If the system rates your SQL depth as strong, it points to the moment in the transcript where you reasoned through an index versus a schema change. Nothing is an opaque number floating on its own.

The recording is yours, and a human can review it

Your interview is recorded and transcribed, and you can read the full transcript. Under the EU AI Act you also have the right to ask a person to review an automated assessment that affects you. That human-review path is built in, not bolted on. If a decision feels wrong, you can challenge it with the same evidence the system used.

Fairness is not a slogan you can verify by reading a marketing page. It is a property you can check by looking at the rubric, the recording, and the route to a human. When all three are visible, a candidate can see exactly why a decision was made and push back if it does not hold up.