Skip to content

Design and implement an MLOps infrastructure Questions

Practice questions for Design and implement an MLOps infrastructure topic in Microsoft Certified: Machine Learning Operations (MLOps) Engineer Associate. 41 questions covering this domain.

41 questions12 easy19 medium10 hard
Q1
medium

A security architect wants jobs to access storage securely without embedding account keys or SAS tokens in code. Which authentication approach is the ...

Q2
hard

A new team member says a datastore and a data asset are the same thing. Which correction is most accurate?

Q3
easy

A practitioner needs a personal development environment for notebooks and experimentation in Azure Machine Learning rather than a shared autoscaling t...

Q4
medium

A team runs intermittent experiments and wants Azure Machine Learning to provision compute only when needed so they do not manage cluster lifecycle di...

Q5
easy

A machine learning team wants one Azure resource that centrally organizes compute, datastores, environments, jobs, endpoints, and models for collabora...

Q6
hard

A regulated company must keep workspace storage access on private network paths and avoid routing traffic over the public internet. Which design is mo...

Q7
easy

A team needs a versioned tabular data asset that can include transformations and be reused across jobs. Which Azure Machine Learning data asset type s...

Q8
medium

An engineer wants infrastructure definitions for workspaces and supporting resources to be repeatable, reviewable, and deployable through source contr...

Q9
medium

Several Azure Machine Learning workspaces need to reuse the same environments, components, and models without recreating them in each workspace. Which...

Q10
medium

A training workload must scale out automatically across multiple nodes when job demand increases and scale back down when idle. Which compute target s...

Q11
easy

When you create an Azure Machine Learning workspace, which dependent Azure resource is automatically provisioned to store secrets such as connection s...

Q12
medium

Compliance requires the workspace and its dependent storage account to encrypt data at rest with a key the security team can rotate and revoke. Which ...

Q13
medium

Developers using personal compute instances often forget to stop them, driving up cost. Which built-in feature should the platform team configure on e...

Q14
medium

An MLOps engineer wants to define an Azure Machine Learning workspace and its dependent resources declaratively so the same configuration can be deplo...

Q15
medium

A DevOps engineer wants a GitHub Actions workflow to authenticate to Azure without storing a long-lived service principal secret. Which approach is re...

Q16
easy

Which Azure Machine Learning asset packages a Docker image plus a Conda specification so jobs run with the same dependencies on every compute target?

Q17
hard

An organization needs a workspace whose data plane (notebooks, jobs, endpoints) is not reachable from the public internet, while still allowing approv...

Q18
hard

An MLOps team must promote the same workspace, compute, and registry topology from dev to prod with reviewable changes and automated validation. Which...

Q19
medium

A team wants reusable, versioned pipeline steps that accept defined inputs and outputs and can be authored once and consumed across multiple pipelines...

Q20
easy

An engineer needs to register a single Parquet file in Azure Machine Learning so a job can reference its exact path as one input. Which data asset typ...

Sign in to see all 41 questions

Create a free account to browse all questions — completely free during our launch phase.