Scorecards & Analytics
Reading Scorecards
How to interpret AI-generated scorecards after a candidate completes an interview.
After a candidate completes their interview, the AI scorer analyzes the transcript and produces a scorecard.
What's on a Scorecard
- Overall Weighted Score — a single number (1-5) based on all competency scores, weighted by your posting configuration
- Recommendation — Advance, Hold, or Decline
- Per-Competency Scores — each competency receives a 1-5 score with:
- Score — numerical rating based on the behavioral anchors
- Evidence — direct quotes from the transcript supporting the score
- Strengths — what the candidate demonstrated well
- Concerns — areas where the candidate was weak or unclear
- Integrity Score — deductions for any integrity flags (tab switching, gaze away, etc.)
Score Scale
| Score | Meaning |
|---|---|
| 5 | Exceptional — clear, specific evidence well above expectations |
| 4 | Strong — solid evidence meeting or exceeding expectations |
| 3 | Adequate — meets basic expectations with some evidence |
| 2 | Below expectations — limited or vague evidence |
| 1 | Insufficient — no meaningful evidence provided |
How Scoring Works
- The scorer never sees the candidate's name or any demographic information
- Scores are based purely on transcript content matched against competency anchors
- The same competency framework is used regardless of interview style (spine or adaptive)
- Each posting's competency weights determine how individual scores combine into the overall score