Wrap-Up Surveys, Aggregated
Every staff member fills out a 4-question survey after each call. In aggregate, those ratings show you where the AI is sharp and where it's still rough.
What we ask
- Did the AI capture the right facts?
- Did the cockpit surface helpful prompts?
- Anything the AI said you'd want to never say again?
- Free-text — what else?
Where to see the results
Settings → Quality, top of page. Aggregate scores by workflow, by tech, by signal. Anything trending down gets a yellow banner.
Acting on it
Drop in fact-capture? Probably a workflow tweak. Wrong prompts surfacing? Tune the signal that fires them. Bad phrasing? Add to forbidden-statements list. Use the survey as a compass, not a leaderboard.