| Fit score | The round’s headline 0–100. Each competency’s answers are averaged, then competencies are weighted by your configured mix and rebased on the weight actually assessed. Bands: 80+ High · 60–79 Medium · 40–59 Low · <40 Reject. |
| Answer score | The 0–10 grade given to a single answer, judged against that question’s sample answer. A score of 0 means answered poorly, not “not asked”. |
| Competency score | The average of all scored answers inside one competency. Question count affects confidence, never weight: two answers and twenty carry the same weight. |
| Rubric | The set of competencies a candidate is scored against, and their weights. See How to Set Scoring Weights. |
| Assessed weight | The share of your competency mix a round actually tested. It is the denominator of the fit score, so a competency nobody asked about never counts against the candidate. |
| Coverage | How much of your configured question list was asked. Reported on every round. On panel and HR rounds it does not reduce the score, the interviewer chose what to ask, not the candidate. |
| Completeness | How much of your competency matrix was assessed: Conclusive, Provisional, Indicative, or Not assessable. It limits what a round may recommend; it never changes the score. |
| Ad-hoc question | A question asked outside your configured list. It counts toward the score when it maps confidently to one of your competencies, and never adds a competency or changes a weight. |
| Evidence | The specific answers behind a score, shown so the number can be checked against what was actually said. |
| Quality review flag | A result from the automated quality checks run on every scored output (schema, score range, citation presence). See AI Observability. |
| Verdict | The recommended outcome for a round, always subject to human review. An Indicative round can only recommend Hold. It can neither advance nor reject. |
| Consolidated score | A single comparable score combining multiple panel inputs. |
| Score override | A reviewer’s replacement for an AI score, always kept with a written justification. |
| Decision reason | The stated reason a reviewer must give whenever they change a candidate’s Decision. |
| Join-likelihood signal | An intelligence read on how likely a candidate is to accept and join. Upcoming. |
