Home · Features · AI CV formatting and candidate matching · Candidate match score
FeatureThree separate AI judges score each CV against the job description, recruiters confirm what is unclear, and every candidate on a role is scored without anyone clicking.
A match score is only useful if it is consistent and you can explain it. One AI asked once will give a different answer on a different day. The candidate match score is built to be steadier than that, and to get better as the recruiter learns more about the candidate.
When a CV is matched to a role, three separate AI judges each compare it with the job description, requirement by requirement, with the evidence from the CV for each one. Their votes are combined into the Requirements Match percentage you see on the pipeline card. Where the judges disagree, the combined score is more cautious, not more confident.
Because the score is built requirement by requirement, you can see why a candidate scored what they did: which must-haves were met, which were partly met and which were missing. That is what makes it defensible to a client, and useful to you on a screening call.
CVs leave things out. For each requirement the CV does not settle, the card lists it under What to confirm, with a slider and a notes box. After your screening call, set how well the candidate meets it and note what they said. The next time the candidate is scored, the judges use what you confirmed, and the note can appear on the CV as checked with the candidate.
Scores get better with every screening call, instead of being fixed by what happened to be on the CV.
Adding someone to a job starts scoring in the background straight away, with a Scoring pill on their card until it is done. Nobody has to remember to click Match, and every candidate on a role has a score you can sort by. To score a whole list at once, tick them and use Match to job description in the Tools tray.
For the candidates who matter most, Deep match runs a slower, more thorough comparison of the candidate against the role. A second opinion before you put a name in front of a client.
A high score means the CV covers the requirements as written. It does not mean the person will be good at the job, and a low score can hide a candidate whose CV undersells them. Our articles on why experience is a weak predictor and the confident, wrong shortlist explain the limits. The score tells you where to look and what to ask. The recruiter decides, and the score sits beside the fit risks, which tell you what to raise.
The score is also only as good as the job description, which is why the CRM scores how complete the JD is before you search.
One AI judgement can vary from run to run. Three independent judges, combined, give a steadier score, which is why we chose accuracy over the cheaper single call.
Yes. The score is built from each requirement, with what the CV showed for it and what still needs confirming.
The score appears on the formatted CV you choose to send, so the client sees a consistent measure from your firm.
Score the candidates again. Requirements that were reworded start fresh, so old answers never attach to a different question.
Book a demo with a live role and a few CVs, and we will score them for you.