In many contact centers, supervisors spend up to 80 percent of their time listening back to calls, producing a just handful of evaluations per month that may not be representative. It's a small and context-dependent sample that often says little about actual performance.

In this session, Allen Crane, currently Executive Director, Operational Excellence and Strategic Analytics at GEICO, will share how he has sought to change this. He'll walk us through his experiences transforming AI-driven agent evaluation systems, discussing how, in three months, he rolled out a gen AI evaluation framework across contact centers for 15,000 agents. In doing this, Allen enabled millions of calls to be assessed and gave supervisors a holistic view of agent performance.

We'll discuss how this approach can be fine-tuned, mitigating AI hallucinations and maintaining a commitment to keeping humans in the loop.

Attendees will learn:

• How to move from a small, unrepresentative sample to millions of fully evaluated calls and what a representative picture reveals
• Why human-in-the-loop isn't a compromise and what makes AI evaluation trustworthy enough to act on
• How to responsibly tie AI-driven evaluation to business KPIs and staffing decisions