Sign inChoose a certification
Microsoft Certified: Machine Learning Operations (MLOps) Engineer Associate · AI-300

Test your free AI-300 learning scope

Sign in to practice every question included in the free learning topics, review explanations, and save your results.

Sign in to start free test
Practice test sheet with answer checks and a timer

How free practice works

Build confidence before your next step

Your account keeps attempts, review data, and the free study tools connected.

  1. 1

    Sign in or register

    Create a free account before starting a test.

  2. 2

    Answer included questions

    Practice questions from the included free learning topics.

  3. 3

    Review your results

    See explanations and knowledge-domain feedback immediately.

Your free test

Practice the included free learning topics

Each attempt selects from the questions inside the administrator-selected free learning topics.

Includedselected free learning topics
Self-pacedestimated time
Instantanswers and explanations
Sign in to start free test

Public question sample

Preview the question format

See the answer and explanation before starting your own practice test. This sample remains the same while it is available.

Sample questionMultiple choice

An engineer creates an Azure Machine Learning AmlCompute cluster whose unattended training jobs will use its attached managed identity to read an ADLS Gen2 account. Storage networking and job submission permissions are already validated. The organization prohibits storing a client secret or storage key for this workload.

The CLI has authenticated subscription, resource-group and workspace defaults. The compute definition includes a supported VM size:

az ml compute create --name training-cluster \
  --type AmlCompute --size Standard_DS3_v2 \
  --min-instances 0 --max-instances 4 \
  --identity-type <IDENTITY_CHOICE>

Which two supported compute-identity configurations meet the requirement when the necessary storage data access is granted? (Choose 2.)

  1. A.Configure --identity-type None and inject a Service Principal Client Secret via plain-text environment variables at job runtime.
  2. B.Set --identity-type UserAssigned and --user-assigned-identities to the pre-provisioned identity resource ID; grant that identity the required storage data-plane permission.Correct answer
  3. C.Configure --identity-type SystemAssigned on the workspace itself, avoiding the need to configure identities on individual compute targets.
  4. D.Set --identity-type SystemAssigned and grant the resulting cluster identity the required storage data-plane permission on the ADLS Gen2 account.Correct answer
  5. E.Use --identity-type WorkspaceManaged to automatically federate the user's interactive login token to the distributed batch nodes.
Why this is correct
AmlCompute supports system-assigned and user-assigned managed identities. The job must actually select the intended compute identity and its effective storage authorization must permit the data read. Attaching an identity, workspace management rights and registering a datastore are separate operations.
Knowledge domain: Design and implement an MLOps infrastructureLearning topic: Create and manage resources in a Machine Learning workspaceDifficulty: beginner

Subject coverage

Knowledge domains covered in this subject

These weights describe the full subject blueprint. Free-test questions are selected randomly from the available question pool.

Design and implement an MLOps infrastructure19%
Implement machine learning model lifecycle and operations29%
Design and implement a GenAIOps infrastructure24%
Implement generative AI quality assurance and observability14%
Optimize generative AI systems and model performance14%

Ready when you are

Put your current knowledge to the test

Practice the included free learning topics and review explanations immediately.