A fully managed AWS service for building, training, and deploying your own machine learning models. Unlike Bedrock (consuming third-party foundation models via API), SageMaker gives you control over the training code, framework, and data, while AWS manages the underlying hardware.
In AIF-C01 (Domain 1) it is compared to Bedrock on the IaaS→Serverless scale: SageMaker is similar to EC2 (more control, more customer responsibility), Bedrock is similar to Lambda (serverless, AWS manages the model). Typical question: training a custom ML model with specific data and framework → SageMaker.
Amazon SageMaker is a concept that can show up on the exam. Check the full guide to see it in context.