AI-300 Practice Test 13 — 15 Questions

Question 1

Which TWO resources are created inside an Azure Machine Learning workspace?

Accepted Answer

Compute cluster.. Option B (Compute cluster) is correct because an Azure Machine Learning workspace hosts managed compute resources such as compute clusters (AmlCompute), which are used to run training pipelines and jobs and are created as child resources of the workspace. Option E (Datastore) is correct because a datastore is a workspace-level resource that stores connection information to Azure storage services (Blob, File, Data Lake, SQL) so experiments can access data without embedding credentials. Option A (External website) is not a workspace resource; a workspace does not create or manage public websites. Option C (Microsoft Entra ID) is incorrect because Microsoft Entra ID (Azure AD) is a tenant-level identity service that exists independently and is only associated with the workspace for authentication, not created inside it. Option D (Office 365 tenant) is incorrect because an Office 365 tenant is a separate Microsoft 365 organizational boundary unrelated to Azure Machine Learning workspace resources.

Answer

External website.

Answer

Azure Active Directory.

Answer

Office 365 tenant.

Question 2

Your team needs to share a model across different workspaces. What is the most efficient way to achieve this in Azure Machine Learning?

Accepted Answer

Use a shared Azure Machine Learning Registry.. Using a shared Azure Machine Learning Registry is the most efficient way to share a model across different workspaces because it provides a centralized, versioned repository for models, components, and environments that can be accessed by multiple workspaces with proper access control. It eliminates duplication and ensures consistency.

Answer

Export the model as a pickle file and upload to each workspace.

Answer

Duplicate the workspace storage account.

Answer

Re-train the model in every workspace.

Question 3

You need to ensure that a training script running on Azure ML has access to secrets (like API keys) without hardcoding them in the script. What should you use?

Accepted Answer

Azure Key Vault. Azure Key Vault is integrated with Azure ML to securely manage and inject secrets into training jobs.

Answer

Environment variables in the Dockerfile

Answer

Local configuration files

Answer

Hardcoding in the script

Question 4

Which TWO types of compute can be used for training in Azure Machine Learning?

Accepted Answer

Compute Cluster.. Compute Cluster (B) is correct because it is Azure Machine Learning's managed, scalable multi-node compute target designed specifically for training and batch scoring jobs, supporting autoscaling and GPU/CPU SKUs. Compute Instance (C) is also correct because it is a managed single-node cloud workstation that can run training scripts and notebooks interactively, making it suitable for development and smaller training workloads. Key Vault (A) is a secrets-management service, not a compute resource, so it cannot train models. Web App (D) is a PaaS hosting service for web applications, and Azure SQL (E) is a managed relational database, neither of which provides the compute environment for ML training.

Answer

Key Vault.

Answer

Web App.

Answer

Azure SQL.

Question 5

What is the primary purpose of a 'Labeling Project' in Azure Machine Learning?

Accepted Answer

To facilitate manual data annotation for supervised learning.. A Labeling Project in Azure Machine Learning is designed to facilitate manual data annotation for supervised learning. It provides a workspace where human labelers (or ML-assisted labeling) tag images, text, or other data to create labeled datasets for training. It does not version datasets, monitor drift, or define evaluation metrics.

Answer

To automatically version datasets.

Answer

To monitor model deployment drift.

Answer

To define model evaluation metrics.

Question 6

You need to orchestrate a series of ML steps (Data prep, Training, Evaluation). Which Azure ML feature is specifically designed for this?

Accepted Answer

Azure ML Pipelines. Azure ML Pipelines allow you to string together components into a workflow.

Answer

Notebooks

Answer

Batch endpoints

Answer

AutoML

Question 7

Which Azure Machine Learning resource provides a pre-configured environment for development?

Accepted Answer

Compute Instance.. A 'Compute Instance' is a managed, cloud-based development environment that comes pre-installed with the Azure ML SDK, Jupyter, and other tools.

Answer

Inference Cluster.

Answer

Compute Cluster.

Answer

Managed Endpoint.

Question 8

What is the 'Workspace' in Azure Machine Learning?

Accepted Answer

The top-level resource for managing all ML assets.. The Workspace is the top-level resource for Azure Machine Learning, providing a centralized place to manage all artifacts, computes, and jobs.

Answer

A specific model deployment.

Answer

A virtual machine for training.

Answer

A storage container for datasets.

Question 9

You are troubleshooting high latency in a RAG-based application. The vector search is fast, but the generation phase is slow. Which component should be scaled?

Accepted Answer

Increase the 'Instances' count for the LLM deployment.. The generation phase is handled by the LLM, not the vector database or embedding model. Increasing the 'Instances' count for the LLM deployment provides more compute capacity to handle concurrent generation requests, directly reducing latency. This is the correct scaling action because the bottleneck is identified as the generation phase, which is compute-bound on the LLM inference servers.

Answer

Optimize the vector database index.

Answer

Reduce the 'temperature' setting.

Answer

Increase the 'Embedding Model' throughput.

Question 10

Which THREE items should you keep track of in your AI Foundry project to ensure good governance?

Accepted Answer

Quota usage.. Quota usage (A) must be tracked because AI Foundry projects consume shared Azure quotas for model deployments, tokens, and compute, and monitoring usage against limits prevents throttling and enables capacity planning. Resource Tags (C) are essential for governance since tags provide the metadata used for cost allocation, ownership attribution, and policy enforcement across the project's associated resources. Role-Based Access Control (RBAC) assignments (D) should be tracked to verify least-privilege access, audit who can manage deployments and data, and detect privilege drift over time. User passwords (B) are not a project governance artifact because identity is handled through Microsoft Entra ID with tokens, MFA, and conditional access rather than password tracking. Personal browser history (E) is unrelated to project governance and raises privacy concerns, so it should never be monitored for this purpose.

Answer

User passwords.

Answer

Personal browser history.

Question 11

You want to evaluate how well your model adheres to specific brand guidelines. Which evaluation method is best suited for this?

Accepted Answer

Custom LLM-as-a-judge evaluation. Custom evaluation using a judge model (LLM-as-a-judge) configured with a rubric allows for checking specific style or brand compliance.

Answer

Token usage threshold alert

Answer

Default coherence metric

Answer

Automated regression testing of code

Question 12

You have a sudden spike in requests. What is the quickest way to check if you have hit your Azure OpenAI token quota?

Accepted Answer

Check the 'Quotas' tab in Azure AI Foundry.. The 'Quotas' tab in Azure AI Foundry provides a real-time view of your current token usage against the allocated quota for Azure OpenAI models. This is the quickest way to check if you have hit your token quota because it directly shows usage and limits without needing to wait for billing cycles or parse logs. It aggregates usage across deployments and models, allowing immediate action.

Answer

Review your monthly invoice.

Answer

Check the Azure Monitor 'Errors' log.

Answer

Check the model endpoint deployment logs.

Question 13

You are setting up an Azure Machine Learning workspace and need to ensure that the workspace is not accessible from the public internet. Which networking feature should you configure?

Accepted Answer

Private Endpoint. A Private Endpoint creates a private IP address within your virtual network for the Azure Machine Learning workspace, effectively removing public internet exposure. It maps to a specific private link resource (e.g., 'amlworkspace') and allows secure access over the Azure backbone network. This is the correct feature to ensure the workspace is not accessible from the public internet.

Answer

Azure Bastion

Answer

Application Gateway

Answer

Service Tags

Question 14

You need to ensure that your LLM responses are filtered for harmful content. Which infrastructure component provides this capability?

Accepted Answer

Azure AI Content Safety. Azure AI Content Safety is the integrated service used within AI Foundry to monitor and filter model outputs.

Answer

Azure Monitor

Answer

Azure Key Vault

Answer

Azure Bastion

Question 15

You are fine-tuning a model on Azure OpenAI and notice the training loss curve is fluctuating significantly. What is the most likely cause?

Accepted Answer

The learning rate is too high.. A fluctuating training loss curve is typically caused by a learning rate that is too high. When the learning rate is too high, the optimization algorithm takes large steps and may overshoot the minimum, causing the loss to oscillate or even diverge. Reducing the learning rate often stabilizes training.

Answer

The model's temperature is set too high.

Answer

The training dataset size is too large.

Answer

The model lacks enough epochs.