What are some components of the training pipeline in SAP AI Core? Note: There are 2 correct answers to this question.

Question

Options

• AInput datasets stored in a hyperscaler object store
• BExecutables that define the training process
• CThe SAP HANA database for model storage
• DAutomated deployment to Kubernetes clusters

• A
  82% (50)
• C
  11% (7)
• D
  7% (4)

Explanation

A and B are correct because the SAP AI Core training pipeline requires two foundational elements: the data (input datasets) and the logic (executables). Input datasets are stored in hyperscaler object stores (such as AWS S3, Azure Blob, or GCS), which AI Core accesses during training runs. Executables define the actual training process - they are the scripts or Docker-based workloads that contain your ML training code and are registered in AI Core.

C is wrong because SAP AI Core does not use SAP HANA for model storage. Trained model artifacts are written back to hyperscaler object stores, not a relational database like HANA.

D is wrong because automated deployment to Kubernetes is not a component of the training pipeline - it describes the serving/inference side of ML workflows. While AI Core runs on Kubernetes internally, that infrastructure is abstracted away from the pipeline definition.

Memory tip: Think of a training pipeline as needing two things - "fuel and instructions". The fuel is your data (option A, in the object store), and the instructions are your executable code (option B). Storage databases and deployment clusters belong to different stages of the ML lifecycle.

No community discussion yet for this question.