review support interaction quality against a scorecard with every score tied to quoted ticket evidence, a sampling plan stating the population, the selection method, and what the sample size can actually support, calibration sessions with reviewer variance reported, findings separated into agent beh
review support interaction quality against a scorecard with every score tied to quoted ticket evidence, a sampling plan stating the population, the selection method, and what the sample size can actually support, calibration sessions with reviewer variance reported, findings separated into agent behavior, missing knowledge, broken process, and product cause, coaching actions with follow-up dates,