You work for a bank. You have been asked to develop an ML model that will support loan application decisions. You need to determine which Vertex AI services to include in the workflow. You want to…

Vertex AI Feature Store is for managing and serving ML features, and Vertex AI Vizier is for hyperparameter tuning; neither directly addresses the need to track epoch-level metrics or compare model versions based on chosen metrics.

Vertex AI Pipelines orchestrates ML workflows, and Vertex AI Vizier is for hyperparameter tuning; neither is the primary service for tracking detailed epoch metrics or comparing model versions, though Pipelines can integrate with tracking services.

Vertex ML Metadata tracks training parameters and lineage of ML artifacts, while Vertex AI Experiments allows for organizing, tracking, and comparing different model training runs and their metrics. Vertex AI TensorBoard is crucial for visualizing detailed metrics per training epoch and comparing training performance across various model versions.

Vertex AI Pipelines orchestrates workflows, and Vertex AI Feature Store manages features, which are not directly used for tracking training parameters, epoch-level metrics, or comparing model versions.