You work at a bank. You have a custom tabular ML model that was provided by the bank's vendor. The training data is not available due to its sensitivity. The model is packaged as a Vertex AI Model ser

Vertex AI Model Monitoring relies on request-response logging and metadata about input features to monitor feature distribution and drift. Uploading the model to the Model Registry and deploying it to an endpoint with logging enabled allows the monitoring service to automatically extract features from the input string for analysis, requiring minimal changes to the vendor container.

Configuring `predict_instance_schema_uri` and `predict_input_drift_threshold` are components of setting up model monitoring, but option A describes the overarching process including deployment and implicitly the necessary logging for monitoring.

Refactoring the serving container to accept key-value pairs would require modifying the vendor-provided code, which contradicts the goal of 'minimal effort' and may not be feasible.

Refactoring the serving container to accept key-value pairs would require modifying the vendor-provided code, which contradicts the goal of 'minimal effort' and may not be feasible.