Missio Systems public knowledge
Model monitoring
Ongoing observation of an AI model and its operating context to detect performance, risk, control, or data changes that require action.
Last reviewed:
Missio definition
Model monitoring is the ongoing observation of performance, inputs, outputs, incidents, control effectiveness, and environmental change after deployment. It should detect both technical degradation and changes in use, population, policy, vendor behavior, or business context.
Monitoring is useful only when thresholds connect to action. Every material signal should have an owner, an escalation route, and a defined response such as investigation, rollback, human review, retraining, restriction, or retirement.
Scope note
This definition supports executive and operating decisions. Applicable law,
contracts, sector rules, and system context may require a narrower definition.