Monitors and assesses the real-world performance and output quality of deployed AI systems. Flags anomalies, conducts prompt and response evaluations, and supports ongoing model quality operations.