Skip to content

Reference: screening results

Use this reference when reading Application Screening Analysis on the vacancy’s applications board. Result availability depends on access, screening capabilities, an active CV Playbook, CV data, and quota.

On-screen label Meaning
CV evidence score The AI reading of the CV as a percentage. It is not a hiring decision.
CV score uses scored criteria only Some criteria are Not evidenced, so the score rests on fewer criteria than the CV Playbook holds.
AI Generated The result has no manual override.
Manual Override A reviewer saved a replacement score. Its value appears farther down the dialog.
CV evidence One row for every CV Playbook criterion the result reviewed.
View Full CV Opens CV details or the document when that data is available.

Under the score, the dialog also counts the CV criteria scored, shows the weighted coverage, and names how many required criteria lack evidence. When the documents evidenced none of the criteria, the dialog shows a sentence instead of a percentage, because nothing was measured.

Field Meaning
Criterion name The requirement from the CV Playbook, marked Required or Preferred.
Score The rating out of the criterion’s own maximum, for example 4 of 5.
Not evidenced The documents do not support a conclusion. It is not the lowest score, a failure or a rejection.
Saved requirement The JD or JS requirement the criterion checks.
Rubric anchor The reviewed anchor the evidence matched.
Passage from the CV The sentence the score was read from, with its Location in the document.
Reading Why the passage supports the score.

A Not evidenced row says whether the CV does not mention the criterion, or mentions related duties only. For related duties, the row shows the passage that was read and labels it as not evidence. Every passage is a statement by the Candidate, not verified evidence.

Below the rows, Saved result configuration lists the criteria version, prompt, recorded model and screening time. Previous results lists earlier results for the same application when they exist.

When a result has an override, the Manual Override panel displays Override Score:. The large number at the top remains the AI score.

The required reason is stored with the result, but the board dialog looks for a different field name. Reason: does not appear on the board. Use the score only after you obtain the override context from your organization’s approved review record.

State Display or action
Screening has not run The Candidate remains in Applied and can show Screen application when every condition passes.
Screening is complete The Candidate appears in Screened with View Application Analysis.
The CV Playbook changed after screening The result is outdated, and Re-Screen can appear.
The Candidate reached a later stage or has interview activity Screening again is unavailable.
The CV came from Upload Here Extracted data is available, but View Full CV has no original file to open.

Screening method weights differ from CV Playbook criterion weights. Criterion weights control the CV result. Method weights combine CV, Zone of Genius (ZoG), and AI interview results into one screening score.

Method Editor label Default Score source
CV CV screening 100% The CV screening score.
ZoG Zone of Genius 0% A completed ZoG Profile Match.
Interview AI interview 0% A completed AI interview score.

Every weight is an integer from 0 to 100. The three values must total exactly 100. A zero-weight method contributes no points and does not block the combined score. Every method above zero must have a result before the combined score is available.

Set these values in the vacancy editor under Screening method weights. The board summary and Screening score card remain in English during an Indonesian session.

Last reviewed
Content owner
Aivident maintainers
Review cadence
Every 180 days

Is something on this page unclear?

Tell public support by email. Include the page address and the part that confused you. Contact public support

Do not submit Candidate data, invitation links, recordings, transcripts, passwords, tokens, or secrets.