Overall evaluation summary and decision flow

Two "overall" views

CVViZ surfaces two separate overall summaries on the candidate profile:

1. Aggregated interviewer feedback — every panelist's submitted scorecard, listed on the Interview Feedback tab.
2. AI Overall Assessment — a computed view aggregating Skills / Experience / Education AI assessments. Display-only.

Aggregated interviewer feedback

On the candidate's Interview Feedback tab, each submission shows as a collapsible panel:

• Panel header: "Feedback added by {user.name} {timestamp}".
• The Overall Rating is surfaced via a star icon next to the header (extracted from the criterion where feedback_type === 'O' && answer_type === 'R').
• Expanding the panel reveals the component with all criteria the interviewer rated and any remarks.

Open multiple panels to compare interviewers side by side.

AI Overall Assessment

The component shows a heading Overall Assessment with the AI's combined score across Skills, Experience, and Education. See Skills, Experience, and Education assessments (AI-driven) for what feeds it.

This is computed automatically when AI evaluation is enabled (Settings → Automations) — there's no manual override or input here.

Reading the two together

The two are independent signals:

• The AI Overall Assessment is paper-fit (CV vs. JD).
• The Aggregated interviewer feedback is human read after meeting them.

Disagreement between the two is informative — strong AI score with weak interview feedback often means the JD/CV match overstates the role fit; weak AI score with strong interview feedback often means a non-traditional CV.

Decision flow

The decision itself — to advance, reject, extend an offer — happens via a status change on the candidate. See Moving candidates through stages and Rejecting candidates.

To capture rationale alongside the status change, use the Remarks field on the Change Candidate Status form (rich-text). That's the canonical place CVViZ stores decision context.

Best practices

• Have everyone submit before the debrief conversation. Reduces anchoring during the discussion.
• Capture the final decision in the Remarks of the status change. Future-you will thank present-you for a one-line rationale.
• Calibrate quarterly. Pull recent candidates' feedback panels, check that ratings track outcomes (e.g. did 5-star hires actually succeed in role?).
• Treat AI Overall Assessment as a hint, not a decision. The interview feedback panels are where the actual decision should live.