MLA-C01 Question #41: Real Exam Question with Answer & Explanation

Question

An ML engineer needs to use AWS CloudFormation to create an ML model that an Amazon SageMaker endpoint will host. Which resource should the ML engineer declare in the CloudFormation template to meet this requirement?

Options

• AAWS::SageMaker::Model
• BAWS::SageMaker::Endpoint
• CAWS::SageMaker::NotebookInstance
• DAWS::SageMaker::Pipeline

Explanation

AWS::SageMaker::Model is correct because it defines the ML model artifact - specifying the container image, model data location (S3 URI), and IAM role - which is the prerequisite resource that a SageMaker endpoint uses to serve predictions. Without declaring the Model resource first, there is nothing for an endpoint to host.

Why the distractors are wrong:

• B (Endpoint): This deploys and serves a model, but it depends on a Model resource already existing - it doesn't create the model itself.
• C (NotebookInstance): This provisions a Jupyter notebook environment for development/experimentation, not for hosting a model in production.
• D (Pipeline): This defines an ML workflow/orchestration pipeline (e.g., training, processing steps), not a hosted inference endpoint.

Memory tip: Think of the three-step SageMaker hosting chain - Model → EndpointConfig → Endpoint. The question asks what creates the model, so the answer is always the first link in that chain: AWS::SageMaker::Model.

No community discussion yet for this question.