Data scientists require fine-grained control over ML workflows, the ability to visualize jobs and workflows as a DAG, and a persistent history of experiments with model lineage for audit and compliance. Which AWS solution meets these needs?

SageMaker Pipelines is designed for building, automating, and managing end-to-end machine learning workflows, offering fine-grained control and visualization as a Directed Acyclic Graph (DAG) within SageMaker Studio. SageMaker ML Lineage Tracking automatically records the steps and artifacts of ML workflows, providing a persistent history of experiments and model lineage crucial for audit and compliance. Incorrect options: AWS CodePipeline is a continuous delivery service, not specifically designed for ML workflow orchestration with DAG visualization or fine-grained ML-specific controls. SageMaker Experiments is primarily for tracking and comparing ML experiments, not for orchestrating entire workflows or providing the comprehensive lineage tracking needed for audit and compliance across a full pipeline.