Question

A financial services company deployed a credit-risk ML model to a SageMaker real-time endpoint six months ago. Compliance auditors now require automated, ongoing detection of when the statistical properties of the live inference data diverge from the data the model was trained on, so the team can be alerted and retrain before predictions degrade. Which AWS capability best meets this governance requirement?

Accepted Answer

Amazon SageMaker Model Monitor with a data-quality monitoring schedule that compares live traffic against a baseline

Answer

AWS CloudTrail logs of every InvokeEndpoint API call for audit review

Answer

Amazon Macie scans of the S3 bucket that stores the training dataset

Answer

SageMaker Model Cards documenting the model's intended use and training details