Real-world data and conditions can drift over time: What is the main reason ongoing monitoring is needed
Data and conditions drift after deployment, so continuous monitoring and retraining are needed to keep performance from degrading.
The question
After deploying a demand-forecasting model, the team sets up continuous monitoring and a maintenance schedule. What is the main reason ongoing monitoring is needed rather than a single pre-release test?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Monitoring is required only to detect server outages, since a validated model's accuracy does not change once it is deployed.Plausible because uptime matters, but it wrongly assumes model accuracy is static after release.
- Real-world data and conditions can drift over time, so performance degrades unless the model is monitored and periodically retrained. ✓Correct: input distributions and conditions change after deployment, so monitoring and retraining are needed to preserve performance.
- Continuous monitoring exists mainly to collect usage analytics that help the business justify the original investment made in the project.Plausible since usage is tracked, but the governance reason is performance drift, not cost justification.
- Scheduled maintenance is only about applying security patches and has no bearing on the model's predictive performance.Plausible as patching matters, but maintenance also covers retraining that sustains predictive quality.
The trap
Assuming a validated model's accuracy stays constant, so monitoring is only about uptime or usage rather than drift. How to remember it
Data and conditions drift after deployment, so continuous monitoring and retraining are needed to keep performance from degrading.
How many of these would you get right?
One of 499 AIGP questions on Certsqill. Take a free five-minute check and see your score per domain — not one number, but which section to open tonight.
Test your AIGP readiness — freeMore Understanding How to Govern AI Development questions
- Recording data lineage and provenance: Which practice would have prevented this gap? →
- It fixes the problem: Why does defining the business context matter before feature engineering? →
- Assess and document the lawful rights to collect and use: Which action most directly addresses the threshold →
- All 135 Understanding How to Govern AI Development questions →
Part of the Certsqill AIGP question bank · Understanding How to Govern AI Development ·
Every answer, right and wrong, comes with its own explanation.
