Question

A financial services company must comply with an internal governance policy requiring that every production ML model has documented intended use, training details, evaluation results, and ethical considerations in a single, versioned, auditable artifact. Compliance auditors need to review this documentation without accessing the SageMaker training or inference infrastructure directly. Which AWS capability best meets this requirement with the least custom development?

Accepted Answer

Create an Amazon SageMaker Model Card for each model to capture intended use, training and evaluation details, and risk considerations in a versioned record that can be exported and shared.

Answer

Store free-form Markdown documents describing each model in an S3 bucket and grant auditors read-only access through a bucket policy.

Answer

Enable SageMaker Model Monitor to continuously capture the model's documentation and export it to CloudWatch Logs for auditor review.

Answer

Use the SageMaker Model Registry approval status field to record intended use and evaluation notes for each model version.