Model Monitoring

Model monitoring is the practice of continuously watching a deployed AI system's behavior to detect problems before they cause damage. It tracks performance metrics, input data distributions, output patterns, and operational signals — looking for drift (the real world changing in ways the model wasn't trained on), data quality failures, accuracy degradation, fairness shifts, or anomalous outputs. Without monitoring, a model that worked at launch can silently become unreliable as customer behavior, market conditions, or upstream data sources change. By the time problems surface through complaints or errors, the model may have been making poor decisions for weeks or months.

Deploying without monitoring is not a technology gap — it's a governance gap. Regulators increasingly expect evidence that AI systems are tracked after launch, with defined thresholds and documented responses when something goes wrong. An organization that cannot answer "how do you know your model is still working?" has no meaningful oversight of the AI making decisions on its behalf. The monitoring plan should exist before deployment, not after the first incident.