METRICRM
FeaturesIndustriesAIPricingResources
//
Talk to us
METRICRM
ProductPricingFeaturesAll modulesIndustriesAIIntegrations
ResourcesUser GuideContact
LegalTerms of ServicePrivacy NoticePDPA NoticeData Processing AgreementCookie Policy
© 2026 Webbalances Solution (003096168-D). All rights reserved.Version 4.62.1
All guidesMetriQABuild a QA scorecardHow AI scores every conversationReview and override QA scoresSampling rules and calibrationRead the quality dashboardCoach with sessions, self-reviews and goalsForecast volume and build the scheduleMy schedule on your phone
User Guide › MetriQA

Review and override QA scores

Check the AI's scores against the conversation, correct them, comment and use the review queue.

For: Supervisors and QA leads

Steps

  1. 1Go to MetriQA → Reviews and open a conversation. Filter To review for conversations a sampling rule queued for you.
  2. 2Read the transcript on the left: the AI's evidence is highlighted in yellow and risk quotes in red. Select a quote on the right to jump to it.
  3. 3Under Your review, change any criterion score the AI got wrong (Pass / Fail for auto-fail criteria, or N/A) and add a reason.
  4. 4Select Save review. Your scores become the conversation's score; the AI's scores are kept, and the change is recorded in the audit log.
  5. 5Add a comment for the agent, or use Add to review queue to come back to it later.

Good to know

  • You cannot review a conversation you handled yourself.
  • If the AI re-scores a reviewed conversation, your score stays.

Related guides

  • Build a QA scorecardDefine what a good conversation looks like: criteria, weights, auto-fail and instructions for the AI evaluator.
  • How AI scores every conversationEvery closed case, email, chat, call and AI-agent conversation is scored automatically, with evidence quotes and confidence.
  • Sampling rules and calibrationRoute a share of AI-scored conversations to human review, and check that reviewers and the AI score alike.
  • Read the quality dashboardAverage score, coverage, auto-fails, scores by agent, AI agent, team and channel, the criteria heatmap and risky conversations.

Still need help? Contact support

On this pageStepsGood to knowRelated guides