Better context.
More meaningful coaching.
Understand what happened, find the moments that matter, and prepare your next conversation with the agent.
Conversation review
Review the evidence. Shape the coaching.
Knowledge & standards
Grounded in your supplied materials. Interpreted with context and confirmed by people.
Six source documents
How to use the findings
- Exact quotes locate evidence; they do not prove an interpretation.
- Holds are measured from recognized timestamps. Missing or ambiguous timing needs review.
- Policy accuracy, booking actions and unseen verification need separate records.
- No official QA score, automatic critical failure or disciplinary decision is produced.
Orientation and the detailed scorecard differ on administration critical classification. General phrasing guidance and scenario scripts can also differ. Ask QA to resolve applicable rules.
Important distinctions
Agent hold: update within five minutes, including the final refresh. Customer inactivity: the scripts describe a check after two minutes, then three further minutes.
The November 1 summary-idle clarification is a recorded user calibration, separate from the original documents. Other criteria still need their own assessment.
Data & access
Cases live in this tab’s memory. Refreshing or closing it clears the workspace. Export a report if you need a copy.
AI review sends the conversation and supplied context to DeepSeek through the protected server API. Provider and hosting retention settings require separate confirmation.
A shared team password protects AI calls. It does not establish an individual audit trail.
Knowledge version: FL-COACH-2026-11-03-v1 · Curated references, not a custom-trained model.