AWS Certified Machine Learning – Specialty — Question 312

A company is setting up a mechanism for data scientists and engineers from different departments to access an Amazon SageMaker Studio domain. Each department has a unique SageMaker Studio domain.

The company wants to build a central proxy application that data scientists and engineers can log in to by using their corporate credentials. The proxy application will authenticate users by using the company's existing Identity provider (IdP). The application will then route users to the appropriate SageMaker Studio domain.

The company plans to maintain a table in Amazon DynamoDB that contains SageMaker domains for each department.

How should the company meet these requirements?

Answer options

• A. Use the SageMaker CreatePresignedDomainUrl API to generate a presigned URL for each domain according to the DynamoDB table. Pass the presigned URL to the proxy application.
• B. Use the SageMaker CreateHumanTaskUi API to generate a UI URL. Pass the URL to the proxy application.
• C. Use the Amazon SageMaker ListHumanTaskUis API to list all UI URLs. Pass the appropriate URL to the DynamoDB table so that the proxy application can use the URL.
• D. Use the SageMaker CreatePresignedNotebooklnstanceUrl API to generate a presigned URL. Pass the presigned URL to the proxy application.

Correct answer: A

Explanation

The CreatePresignedDomainUrl API is the correct choice because it generates a authorized, presigned URL specifically for launching an Amazon SageMaker Studio domain for a given user profile. The CreatePresignedNotebookInstanceUrl API is incorrect because it is used for traditional SageMaker notebook instances rather than SageMaker Studio domains. The CreateHumanTaskUi and ListHumanTaskUis APIs are unrelated to domain access, as they are used for configuring human review workflows in SageMaker Ground Truth.