Rerun the AI summary

An evaluation is a snapshot of what the Assistant knew at the time it ran. When the picture changes — a CV is updated, the candidate replies, your team adds comments — you don't have to spend a fresh token to bring the summary current.

If the candidate's profile was updated after the AI evaluation ran, the evaluation section shows:

"This view is based on earlier data. Run the summary again to include the latest updates."

Choose Rerun summary to trigger a fresh evaluation. This does not consume an additional token.

When it's worth rerunning

• After a candidate answers your clarifying questions or a reply lands in the thread — the new context can move a borderline rating or close a Clarifications needed gap.
• After your team adds scorecard comments or notes you want reflected in the narrative.
• After the candidate's CV or profile is refreshed.

Good to know. Don't see the "based on earlier data" banner? Then nothing material has changed since the last run and there's nothing new to fold in — the current summary already reflects the latest. Rerunning is for catching the evaluation up after the inputs change, not for rolling the dice on a different answer.