In an ML organization, teams are wastefully re-creating the same features separately. Which SageMaker feature lets you store features (feature data) used by multiple teams and models centrally so they can be shared and reused?
Choosing the repository that stores, shares, and reuses features.
Amazon SageMaker Feature Store
Correct. SageMaker Feature Store is a repository that stores features centrally so they can be shared and reused across multiple teams and models. The same features can be used at both training time and inference time.
Amazon SageMaker Autopilot
Autopilot is an AutoML feature that automatically builds models from data.
It is not a repository that stores, shares, and reuses features, so it is incorrect.
Amazon SageMaker Ground Truth
Ground Truth is a data labeling service that applies labels to training data.
It is not a repository that stores, shares, and reuses features, so it is incorrect.
Amazon SageMaker Clarify
Clarify is a feature that provides bias detection and explainability.
It is not a repository that stores, shares, and reuses features, so it is incorrect.
Organize the main SageMaker features by purpose.
・Feature Store: stores features centrally to share and reuse them across multiple teams and models (the correct answer here).
・Autopilot: AutoML that automatically builds models from data.
・Ground Truth: data labeling that applies labels to training data.
・Clarify: provides bias detection and explainability.
・JumpStart: a hub of pretrained models and ready-made solutions.
Even when the names are similar, their roles differ, so map them by 'what the feature does.'