Review AI-extracted insights

Once transcripts are in, you can ask the project questions in plain language rather than guessing at keywords.

Asking

Ask the way you would ask a colleague: what the decision on the finance migration was, or who raised the concern about customer data. The search ranks passages by meaning rather than by matching words, so it finds the exchange where somebody described the problem without using your term for it.

What comes back

Relevant passages with the meeting and the timestamp they came from. That provenance is the point: you are not being told what was decided, you are being taken to where it was said, and you read it yourself.

Scope

Results come only from transcripts on this project. This is a boundary, not a filter you can widen: meetings from another engagement will not appear even if the same attendees are in them. The account-level Intelligence surface is the one that reaches across projects.

What it is for

Two things, mostly. Finding the political problem in the record, the concern somebody raised three weeks ago that has since become the blocker. And settling the what-did-we-agree question with the actual sentence rather than with two recollections.

What it will not do

It will not decide anything, score anything, or summarise your deal into a recommendation. It retrieves and it cites. The reading is yours.