Statuses and advisory scoring
Interpret analysis states, role-fit bands, interview states, recommendations, and confidence.
Analysis state
| State | Meaning | Next action |
|---|---|---|
| Processing | Upload, extraction, or AI analysis is still running. | Wait or return to Reports later. |
| Completed | The report is available for review. | Inspect the evidence and source context. |
| Mismatch review needed | Very weak role evidence paused full publication. | Verify the inputs, then return or proceed explicitly. |
| Failed | The run did not produce a complete report. | Read the error and use the displayed retry path. |
Staged advisory score
Reports use one 0–10 display slot with an explicit source label:
- Resume–JD provisional is shown before usable interview evidence exists. It summarises only job-related evidence from the resume against the supplied job description.
- Overall evidence is shown after a completed interview has enough rated evidence. It combines resume and interview evidence criterion by criterion; it is not a simple average of two headline scores.
The final calculation weights each job-related criterion by importance and by source reliability. Interview evidence has more influence on capabilities that can be demonstrated in conversation, while the resume retains more influence on experience history and scope. Missing or insufficient interview evidence keeps the assessment provisional rather than silently treating missing evidence as a zero.
| Score | Evidence band | Interpretation |
|---|---|---|
| 8–10 | Strong evidence | Strong role-related evidence is present, subject to the stated gaps and source quality. |
| 6–7.9 | Mixed evidence | Meaningful relevant evidence exists alongside important gaps or uncertainty. |
| 3–5.9 | Limited evidence | Some evidence may be relevant, but major requirements are missing or weakly supported. |
| 0–2.9 | Insufficient evidence | The supplied material does not establish enough direct evidence for the role. |
Resume-versus-interview evidence matrix
The matrix makes the final score auditable. For each criterion it shows the resume score, interview score, combined score, evidence, and one of these states:
| State | Meaning |
|---|---|
| Corroborated | The sources are materially consistent. |
| Strengthened | Interview evidence raises confidence in the resume evidence. |
| Weakened | Interview evidence is weaker than the resume evidence. |
| Contradicted | The sources materially conflict and need human review. |
| Resume only / Interview only | Only one source contains usable evidence. |
| Evidence gap / Not assessed | Neither source supports a reliable criterion score. |
Coverage and confidence describe how complete and reliable the source evidence is. They should qualify the score, not be treated as additional candidate attributes.
Findings that do not affect scoring
Two finding classes are intentionally outside the score:
- Data integrity: name, email, profile, resume-document, or document-integrity mismatches. High-priority issues should be resolved before the report is used because the sources may belong to different records.
- Compliance and excluded criteria: protected or otherwise disallowed job-description criteria that were removed from analysis. These explain what was excluded and must not appear as candidate risk or evidence against the candidate.
The Candidate risks and red flags section is reserved for job-related concerns supported by candidate evidence, such as an unverified required capability or a contradiction between candidate sources.
Interview state
| State | Meaning |
|---|---|
| Pending | The invitation exists and the candidate has not started. |
| In progress | The candidate session has started and is not final. |
| Completed | The conversation finished; rating may still process briefly. |
| Expired | The seven-day invitation window has passed. |
Interview recommendation
A completed rating can recommend Proceed, Borderline, or Reject. The output can include readiness, decision confidence, competency scores, evidence coverage, recommendation risk, strengths, and weaknesses.
Decision confidence means confidence in the generated interpretation of available evidence. It is not certainty about candidate truthfulness, future performance, or the correct hiring decision.
Always inspect transcript evidence, question coverage, source quality, role alignment, accessibility considerations, and technical interruptions before acting.