Why it matters
Your QA rubric should reflect what a good call means in your operation. Define the required explanation, disclosure or resolution step, then check whether it happened.
A quoted moment makes the finding easier to inspect. Supervisors can check the context, correct a verdict and keep a record of the decision.
How it works
- 01
Write your criteria
Turn a policy or observable behavior into a check. Add the policy text needed to interpret it.
- 02
Test and publish
Run a draft rubric against a real call, review the results and publish a version for use.
- 03
Review the result
Check the quote and surrounding conversation. Confirm the finding or save a human correction.
What’s included
Rubric editor
Create your own rubrics, test drafts, publish versions and view their history. No model subscription is required.
Rules and model judgments
Use direct checks for clear requirements and semantic judgments for criteria that need context.
Evidence and escalation
Call1 checks that quoted evidence appears in the transcript. Missing evidence or uncertain results can go to human review.
QA trends
Track scores, call volume and criterion outcomes by rubric and date. Open the underlying calls when a pattern needs investigation.
Your data
Rubric scoring uses masked text locally. Your custom policies and transcripts are not automatically added to an industry model.
Before you roll out
Check relevance as well as wording
A quote can be real and still fail to support the verdict. Review both the quoted words and how they relate to the criterion.
Compare reviewer decisions
Use a separate set of calls to compare automated findings with human judgments. Include corrections when measuring assisted review time.
Availability
- Custom rubric authoring and scoring are part of the free platform. Proposed Industry Advantage packs would provide maintained starting content.
- People remain responsible for coaching and personnel decisions. A rubric pack does not establish regulatory compliance.
Questions for your team
- 01
Is each criterion clear enough for reviewers to apply consistently?
- 02
Does every finding have relevant transcript evidence?
- 03
Which errors would change a coaching or policy decision?