QA scorecards and AI call scoring
Score calls against your standards with Dialstitch QA scorecards. The AI drafts every evaluation, reviewers finish it, and agents can acknowledge or dispute.
QA scorecards turn your call standards into a checklist. The AI drafts an evaluation for every answered call, and your reviewers check and submit it. Open Quality in the sidebar. AI QA scorecards and call scoring are part of the Contact Centre plan.
Build a scorecard (admins)
Open Scorecards
Go to Quality, then the Scorecards tab.
Add sections and criteria
Group criteria into sections. Each criterion has a name, guidance for the reviewer and AI, a weight, a Yes/No or 1 to 5 answer, and whether N/A is allowed.
Choose where it applies
Under Applies to, pick numbers, queues or teams. Leave it empty to score every answered call.
Turn it on
Tick Active: draft evaluations for new calls. Changes apply to calls scored after you save.
How evaluations work
- After an answered call is transcribed, the AI drafts an evaluation: an answer, its reasoning and a supporting quote for each criterion.
- A reviewer opens it from the Review queue, edits anything, and selects Save draft or submits it.
- The agent sees it under My evaluations and can Acknowledge or Dispute it. Disputes need a comment.
Owners and admins are reviewers. Members only see their own evaluations, after they're submitted.
Customer satisfaction
Turn on Predict a 1-5 customer satisfaction score for each answered, transcribed call to get an AI CSAT score on every call. It's a prediction from the transcript, not a survey.
Summary
The Summary tab shows quality across your team. Analytics has a Quality tab too.