AI-300 Genaiops Infrastructure Practice Test 4 — 15 Questions

Question 1

You are troubleshooting a failed agent deployment. Which TWO areas should you inspect first?

Accepted Answer

The endpoint availability status.. Option B (the endpoint availability status) is correct because a failed agent deployment often stems from the target endpoint being unreachable, unhealthy, or misconfigured, so checking its availability status quickly reveals connectivity or service-health problems that would block deployment. Option C (the 'Deployment Logs' in the Azure AI Foundry portal) is correct because these logs capture the specific error messages, stack traces, and failure reasons generated during the deployment attempt, making them the most direct source for diagnosing why the agent failed. Options A, D, and E do not belong: the Azure status page only reports broad, platform-wide outages rather than issues specific to your agent deployment, user profile settings are unrelated to deployment execution, and the billing usage report reflects cost consumption rather than deployment health or errors.

Answer

The Azure status page.

Answer

The user profile settings.

Answer

The billing usage report.

Question 2

Your team wants to monitor the safety of model outputs. Which THREE steps should you take?

Accepted Answer

Configure 'Diagnostic Logs' to track flagged content.. Option A is correct because configuring Diagnostic Logs in Azure AI services captures flagged content events and prompt/response metadata, giving the team an auditable trail of unsafe model outputs for review and compliance. Option B is correct because enabling Content Safety filters on the deployment applies the harm-category classifiers (violence, hate, sexual, self-harm) to both prompts and completions, actively blocking or annotating unsafe outputs at inference time. Option C is correct because setting up Alerts on metrics such as content violation rates (via Azure Monitor) notifies the team when unsafe output frequency spikes, enabling rapid incident response. Option D does not belong because bypassing filters removes the very safety controls being monitored and increases risk. Option E does not belong because disabling logging eliminates the telemetry needed to detect, investigate, and alert on unsafe outputs.

Answer

Allow all models to bypass filters for performance.

Answer

Disable all logging to save storage space.

Question 3

You want to automate the scaling of your AI agent orchestration. Which TWO components do you need to configure?

Accepted Answer

Autoscale policies.. Auto-scaling is driven by metrics and defined by policies that govern instance counts.

Answer

A manual human-in-the-loop review.

Answer

Hardcoded instance counts.

Answer

Static IP addresses.

Question 4

To optimize the cost of your AI infrastructure, which THREE actions should you consider?

Accepted Answer

Deleting unused model deployments.. Option A is correct because deleting unused model deployments eliminates the ongoing compute charges (e.g., per-hour inference endpoint or provisioned throughput costs) that accrue even when a deployment receives no traffic, directly reducing waste. Option B is correct because purchasing reserved capacity (such as Azure OpenAI Provisioned Throughput Units or reserved VM instances) commits to a 1- or 3-year term in exchange for a significant discount versus pay-as-you-go pricing, which lowers cost for predictable, long-term workloads. Option D is correct because right-sizing instances based on actual usage metrics (CPU, GPU, memory, and token throughput) ensures you pay only for the capacity you truly need, avoiding over-provisioned and underutilized resources. Option C is not correct because deploying every model in every region multiplies redundant compute and storage costs without guaranteeing usage, which increases rather than optimizes spend. Option E is not correct because always choosing the 'Premium' SKU incurs higher charges for features that may not be required, whereas selecting the appropriate SKU tier for the workload is the cost-efficient approach.

Answer

Deploying all models in every available region.

Answer

Always selecting the 'Premium' SKU.

Question 5

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 6

You are experiencing throttling on your AI endpoint. Which TWO steps should you take?

Accepted Answer

Check current usage against quota limits.. Option A is correct because throttling on an AI endpoint typically occurs when request rate or token consumption exceeds the provisioned quota, so you must first check current usage against quota limits (e.g., requests per minute, tokens per minute) to confirm which limit is being hit. Option D is correct because, once you confirm the workload legitimately exceeds the assigned quota, the standard remediation is to request a quota increase for the model through the service's quota management process. Option B is wrong because disabling security features does not affect throttling and would expose the endpoint to risk. Option C is wrong because swapping to a faster but less capable model changes capability, not the quota or throttling behavior, and is not a proper fix. Option E is wrong because deleting existing deployments would remove capacity and disrupt service rather than resolve throttling.

Answer

Disable all security features.

Answer

Change the model version to one that is faster but less smart.

Answer

Delete all existing deployments.

Question 7

Which THREE features does Azure AI Foundry provide to help developers?

Accepted Answer

Evaluation.. Azure AI Foundry provides Evaluation (A) so developers can systematically assess model outputs using built-in metrics such as groundedness, relevance, fluency, and similarity, which is essential for validating generative AI applications before deployment. Model Catalog (D) is correct because it offers a curated library of foundation models from providers like Azure OpenAI, Meta, Mistral, and Hugging Face, letting developers discover, compare, and deploy models directly within the platform. Prompt Flow (E) is correct because it supplies a development tool for orchestrating LLM, prompt, Python, and tool nodes into executable flows, enabling prototyping, testing, and production deployment of prompt-based apps. Microsoft Entra ID creation (B) is not a Foundry feature—identity and tenant management belong to Microsoft Entra ID (formerly Azure AD) in the Azure portal. Direct physical hardware access (C) is not offered, since Foundry delivers managed, serverless or hosted compute abstractions rather than bare-metal physical access.

Answer

Azure Active Directory creation.

Answer

Direct physical hardware access.

Question 8

Which THREE metrics are critical for monitoring the health of a GenAI deployment?

Accepted Answer

Request Latency.. Request Latency (A) is critical because it measures the end-to-end time for inference requests, directly reflecting user-perceived responsiveness and helping detect model or infrastructure slowdowns in a GenAI deployment. Tokens Per Minute (B) is essential since GenAI workloads are token-based; it quantifies throughput and capacity utilization, revealing whether the deployment is keeping up with demand or hitting rate limits. Error Rate (E) must be monitored to catch failed inferences, timeouts, throttling (HTTP 429), and model or API errors that degrade reliability. User email addresses (C) are personally identifiable information, not a health or performance metric, so they are irrelevant to monitoring. CPU Usage of the host node (D) is a generic infrastructure metric and, while sometimes useful, is not one of the critical GenAI-specific health metrics compared to latency, token throughput, and error rate.

Answer

User email addresses.

Answer

CPU Usage of the host node.

Question 9

Which TWO of the following are benefits of using Prompt Flow for orchestration?

Accepted Answer

Integrated testing and evaluation.. Option B is correct because Prompt Flow provides built-in tooling for testing and evaluating flows, including the ability to run batch evaluations against datasets and compute metrics such as groundedness, relevance, and coherence, which supports iterative quality improvement. Option E is correct because Prompt Flow offers a visual designer (in Azure AI Foundry / Azure Machine Learning) where you can author and connect nodes—LLM, Python, prompt, and tool nodes—into directed acyclic graphs, making complex orchestration easier to design and understand. The unmarked options do not belong: A is wrong because Prompt Flow orchestrates calls to LLMs rather than removing the need for one, C is wrong because authentication and authorization are still required for accessing models and connected resources, and D is wrong because Prompt Flow does not automatically manage hardware—compute provisioning and scaling remain the user's responsibility.

Answer

Removal of the need for an LLM.

Answer

No need for authentication.

Answer

Automatic hardware management.

Question 10

A team is scaling a GenAI application that utilizes multiple models via model-as-a-service. You notice that inference requests are being throttled. What is the most effective infrastructure-level adjustment to handle the increased load?

Accepted Answer

Purchase and deploy Provisioned Throughput Units (PTUs) for the specific model deployment.. Provisioned throughput units (PTUs) allow for dedicated capacity, which is the standard way to prevent throttling for high-scale GenAI workloads in Azure.

Answer

Change the model version to a lower-parameter variant.

Answer

Move the deployment to a different subscription within the same Azure region.

Answer

Increase the number of instances in the associated Azure Kubernetes Service (AKS) cluster.

Question 11

You are setting up a CI/CD pipeline for GenAI models. You need to evaluate the model's performance on a benchmark dataset before it is promoted to the production endpoint. Which Azure AI Foundry capability should be integrated into your pipeline?

Accepted Answer

Prompt Flow evaluation runs.. Evaluation runs in Prompt Flow allow for automated testing of model quality against datasets, which is a required step for a robust GenAI CI/CD pipeline.

Answer

Model catalog benchmarking.

Answer

Azure AI Content Safety scanning.

Answer

Azure Machine Learning model drift monitor.

Question 12

You are designing the infrastructure for a multi-agent system. Which THREE of the following are essential components to ensure agent orchestration scalability?

Accepted Answer

Horizontal Pod Autoscaling (HPA) for inference services.. Option B is correct because Horizontal Pod Autoscaling (HPA) dynamically adjusts the number of pod replicas for inference services based on metrics like CPU utilization or custom metrics, which is essential for scaling agent workloads elastically as demand fluctuates. Option C is correct because a message queue such as Azure Service Bus decouples agent tasks, enabling asynchronous communication, load leveling, and reliable task distribution across multiple agents, which is fundamental to orchestration scalability. Option D is correct because a distributed cache like Azure Cache for Redis provides low-latency shared state management, allowing agents to coordinate, share session context, and avoid bottlenecks from repeated database reads in a scaled-out system. Option A is not essential for orchestration scalability, since Azure Data Lake Storage is a storage solution for large-scale data and model artifacts, not a runtime scaling or coordination component. Option E is not essential either, as Azure Bastion provides secure RDP/SSH access to virtual machines for management purposes and does not contribute to agent orchestration scalability.

Answer

Azure Data Lake Storage for model weights.

Answer

Azure Bastion for secure management.

Question 13

A developer is troubleshooting an agent that is failing to connect to an external tool. Which THREE of the following infrastructure settings should be verified?

Accepted Answer

The Azure AI Foundry 'Connections' configuration for the specific tool.. Option A is correct because the Azure AI Foundry 'Connections' configuration defines the endpoint, authentication, and credentials the agent uses to reach an external tool, so a misconfigured connection would directly cause connection failures. Option B is correct because the Prompt Flow runtime relies on environment variables (for example API keys, endpoint URLs, or proxy settings) to authenticate and route requests to external tools, and missing or incorrect values break connectivity. Option E is correct because outbound traffic from the runtime must be permitted by Network Security Group rules; if the NSG blocks the destination port or IP, the agent cannot establish the connection. Option C is not relevant because the Azure Resource Group location is a metadata/placement attribute and does not govern outbound connectivity or tool authentication. Option D is not relevant because the LLM model version affects model behavior and output quality, not the network or credential path used to connect to an external tool.

Answer

The Azure Resource Group location.

Answer

The model version of the LLM.

Question 14

Your team runs a GenAI agent in Azure AI Foundry that calls an Azure OpenAI model deployment. The agent intermittently fails with authentication errors, and you suspect the managed identity used by the agent's compute does not have the required role assignment on the Azure AI Foundry resource. You need to verify and fix the role assignment. Which action should you take?

Accepted Answer

Assign the Cognitive Services OpenAI User role to the agent's managed identity at the scope of the Azure AI Foundry resource or the specific model deployment.. Authentication failures from a managed identity calling a model deployment typically stem from a missing data-plane role assignment. The Cognitive Services OpenAI User role grants exactly the permission required to invoke deployments, and assigning it to the identity at the Foundry resource or deployment scope resolves the error while following least privilege.

Answer

Add the agent's managed identity to the Contributor role on the resource group that contains the Azure AI Foundry resource.

Answer

Regenerate the Azure AI Foundry project's API key and update the agent's connection string to use the new key.

Answer

Enable public network access on the Azure AI Foundry resource so the agent's compute can reach the endpoint.

Question 15

You manage a GenAI application in Azure AI Foundry that uses a provisioned throughput deployment for a large language model. Finance asks you to reduce cost during off-peak hours without affecting peak-hour performance. The deployment currently runs at a fixed provisioned capacity around the clock. Which approach should you use?

Accepted Answer

Create a second pay-as-you-go deployment and route off-peak traffic to it while scaling the provisioned deployment down during off-peak hours.. Provisioned throughput is billed for reserved capacity regardless of utilization, so the savings must come from not reserving that capacity off-peak. Routing off-peak traffic to a pay-as-you-go deployment and scaling the provisioned deployment down keeps peak performance intact while cutting cost when demand is low. This hybrid routing is the intended pattern for variable demand.

Answer

Enable autoscaling on the provisioned deployment so it scales to zero when there is no traffic.

Answer

Move the model to a batch deployment and process all off-peak requests through the batch API.

Answer

Lower the provisioned throughput units permanently to the minimum supported value and rely on the model's built-in burst capacity during peak hours.