AI-102 Plan and manage an Azure AI solution Practice Test 5 — 15 Questions

Question 1

Refer to the exhibit. You are configuring content filtering for an Azure OpenAI deployment using a JSON policy. You want to block all content with a 'Hate' severity of 'medium' or higher. What should you modify in the policy?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Modify the "severityThreshold" for "Hate" to "medium".. In Azure OpenAI content filtering, the 'severityThreshold' for a given category (like 'Hate') defines the minimum severity level at which content is filtered. Setting it to 'medium' means all content with a severity of 'medium' or higher (including 'high') will be blocked, which matches the requirement to block 'Hate' severity 'medium' or higher.

Answer

Change the "severityThreshold" for "SelfHarm" to "low".

Answer

Change the "contentTypes" to include only "Normal".

Answer

Add a "policyAction" with "block" to the "BlockHateSpeech" policy.

Question 2

Refer to the exhibit. You are creating an Azure AI Search index. You want to enable semantic search using the 'content' field as both the title and the content. However, you receive an error that semantic search cannot be configured. What is the most likely reason?

Accepted Answer

The 'contentFields' list is empty; semantic search requires at least one content field or keywords field.. Semantic search in Azure AI Search requires at least one field specified in the 'contentFields' list to provide the textual content for semantic ranking. If the 'contentFields' list is empty, the service cannot extract semantic meaning from the index, resulting in the configuration error. The 'content' field can serve as both title and content if properly assigned, but the error indicates the list itself is missing entries.

Answer

The 'content' field is of type 'Edm.String', which is not supported for semantic search.

Answer

The 'content' field is not searchable.

Answer

The index does not have a sortable field for the title.

Question 3

Refer to the exhibit. You deployed a Conversational Language Understanding (CLU) project. A user says 'hi, where is my order?' The model returns a single intent 'Greeting' with confidence 0.85. You need the model to detect both intents. What should you change?

Accepted Answer

Set 'multipleIntents' to true.. The Conversational Language Understanding (CLU) service supports multi-intent detection only when the 'multipleIntents' flag is explicitly set to true in the project settings. By default, CLU returns only the highest-confidence intent; enabling this flag allows the model to return all intents whose confidence exceeds the threshold, enabling detection of both 'Greeting' and 'OrderStatus' from a single utterance.

Answer

Lower the confidenceThreshold to 0.5.

Answer

Add more utterances to the 'OrderStatus' intent.

Answer

Change the project kind to 'Orchestration'.

Question 4

Your team is building an AI-powered search application using Azure AI Search with semantic ranking. During testing, the search results are not showing the expected relevance improvements. What should you check first?

Accepted Answer

Confirm that the search service is at least Standard S1 and semantic ranking is enabled. Semantic ranking in Azure AI Search requires a service tier of at least Standard S1 (or higher) and the feature must be explicitly enabled on the index. Without meeting these prerequisites, the search results will not benefit from semantic reranking, regardless of other configurations. Therefore, confirming the service tier and semantic ranking enablement is the first and most critical check.

Answer

Verify that CORS is enabled for your application

Answer

Check that your index has an RBAC role assignment for semantic search

Answer

Ensure the search service is on the Free tier

Question 5

Which TWO actions are recommended to secure an Azure AI Language resource?

Accepted Answer

Use managed identities for authentication from Azure services. Option A is correct because using managed identities lets Azure services authenticate to the Azure AI Language resource through Microsoft Entra ID without storing or rotating access keys, eliminating a common credential-leak risk. Option C is correct because disabling public network access and configuring a private endpoint restricts traffic to your virtual network via Azure Private Link, removing exposure to the public internet. Option B is not recommended because local authentication with access keys relies on shared secrets that can be leaked or over-shared, whereas keyless Entra ID auth is preferred. Option D is wrong because allowing all Azure services through the firewall broadly opens the resource rather than limiting it to trusted networks. Option E is wrong because anonymous access would let unauthenticated callers use the API, which is the opposite of securing the resource.

Answer

Enable local authentication with access keys

Answer

Allow all Azure services through the firewall

Answer

Enable anonymous access for public APIs

Question 6

You see the exhibit in an Azure AI Search skillset definition. The custom skill is not executing as expected. What is the most likely cause?

Accepted Answer

The skill is missing the 'uri' property to specify the endpoint. The custom skill is not executing because the 'uri' property, which specifies the endpoint URL for the web API, is missing from the skill definition. Azure AI Search requires this property to route requests to the external service; without it, the skill cannot be invoked and will fail silently or return errors.

Answer

The input source path is incorrect

Answer

The JSON is malformed and will not parse

Answer

The type should be 'Microsoft.Skills.Custom.WebApiSkill' with dots

Question 7

You are deploying an Azure AI Language service custom text classification model. After training, the model achieves 95% accuracy on the test set but only 60% on a held-out validation set. What is the most likely cause?

Accepted Answer

Overfitting to the training data. A 95% accuracy on the test set but only 60% on a held-out validation set is a classic sign of overfitting. The model has memorized patterns specific to the training and test sets (which may share distribution or preprocessing artifacts) but fails to generalize to unseen data. In Azure AI Language custom text classification, overfitting often occurs when the model is too complex relative to the amount of training data or when hyperparameters like learning rate or number of epochs are not tuned properly.

Answer

Data leakage between training and test sets

Answer

Insufficient training data

Answer

Label imbalance in the training data

Question 8

You need to implement content moderation for a social media platform using Azure AI Content Safety. The solution must block hate speech and self-harm content while allowing mild profanity. Which configuration should you use?

Accepted Answer

Enable hate speech and self-harm with severity threshold 4; enable profanity with severity threshold 6. Azure AI Content Safety allows you to set severity thresholds per category (0-6). By enabling hate speech and self-harm with a threshold of 4, you block content at severity level 4 or higher (e.g., severe hate speech and self-harm), while enabling profanity with a threshold of 6 means only the most extreme profanity (level 6) is blocked, allowing mild profanity through. This matches the requirement to block hate speech and self-harm but allow mild profanity.

Answer

Enable all categories with severity threshold 0 (block all)

Answer

Enable all categories with severity threshold 2

Answer

Enable hate speech and self-harm categories; disable profanity

Question 9

You are deploying an Azure AI Speech custom voice model. After training, the voice sounds unnatural. Which parameter should you adjust to improve naturalness?

Accepted Answer

Use a different language for the training data. Using a different language for the training data (Option D) is necessary when the training language does not match the target voice language, leading to unnatural prosody and pronunciation. Azure AI Speech custom neural voice requires the training data to be in the same language as the desired voice. If a mismatch occurs, switching to the correct language data will significantly improve naturalness.

Answer

Change the audio output format to 48 kHz

Answer

Increase the speaking rate

Answer

Increase the training duration

Question 10

Which TWO of the following are valid ways to authenticate to Azure AI services?

Accepted Answer

Subscription key (API key). Option A (Subscription key / API key) is correct because Azure AI services (such as Azure OpenAI, Computer Vision, and Language) accept a resource-specific key passed in the Ocp-Apim-Subscription-Key header or api-key header for authentication. Option D (Managed Identity with Microsoft Entra ID) is correct because Azure AI services support Microsoft Entra ID (formerly Azure AD) token-based authentication, and a managed identity assigned to a compute resource (e.g., VM, App Service, AKS) can obtain an access token for the https://cognitiveservices.azure.com/ scope without storing secrets. Option B is not valid because SAS tokens are used for Azure Storage data-plane access, not for authenticating to Azure AI services. Option C is not valid because the OAuth 2.0 client credentials flow is a general Microsoft Entra ID token-acquisition mechanism, not a distinct authentication method offered by Azure AI services; the service itself only accepts Entra ID tokens (which may be obtained via various flows) or keys. Option E is not valid because X.509 certificates are not a supported authentication mechanism for Azure AI services.

Answer

Shared access signature (SAS) token

Answer

OAuth 2.0 client credentials flow with Microsoft Entra ID

Answer

X.509 certificate

Question 11

You are configuring semantic search in Azure AI Search. Based on the exhibit, which field is used as the title field for semantic ranking?

Accepted Answer

title. In Azure AI Search semantic ranking, the 'title' field is designated as the primary field for semantic ranking. This field provides a concise, high-level summary of the document content, which the semantic ranker uses to understand the document's main topic and improve relevance scoring. The semantic configuration references this field via the 'titleField' parameter in the semantic configuration object.

Answer

my-semantic-config

Answer

description

Answer

prioritizedKeywordsFields

Question 12

You are designing an AI solution that uses Azure OpenAI Service to answer customer queries. The solution must ensure that the model does not generate harmful or inappropriate content. Which Azure AI service should you configure to enforce content safety policies?

Accepted Answer

Azure AI Content Safety. Azure AI Content Safety is the dedicated service for detecting and filtering harmful content such as hate speech, violence, self-harm, and sexual content. It provides configurable severity thresholds and can be integrated with Azure OpenAI Service to enforce content safety policies on model inputs and outputs. This ensures the AI solution meets responsible AI requirements without modifying the underlying model.

Answer

Azure AI Speech

Answer

Azure OpenAI Service

Answer

Azure AI Search

Question 13

You are planning an Azure AI solution that uses Azure AI Document Intelligence to extract data from invoices. The solution must handle large volumes of documents with varying layouts. Which pricing tier should you choose to ensure scalability and high throughput?

Accepted Answer

Standard (S0). The Standard (S0) tier is correct because Azure AI Document Intelligence requires the S0 tier for production workloads that demand high throughput and scalability across varying document layouts. The S0 tier supports unlimited transactions per second (TPS) with auto-scaling, whereas lower tiers impose strict rate limits or lack the capacity to handle large volumes of invoices with diverse formats.

Answer

Premium (P0)

Answer

Free (F0)

Answer

Basic (B0)

Question 14

You are reviewing the ARM template snippet for an Azure AI Foundry hub. After deployment, you notice that the hub cannot connect to the storage account. What is the most likely cause?

Accepted Answer

The hub uses a managed network that blocks outbound traffic to the storage account. The most likely cause is that the hub uses a managed network which, by default, blocks outbound traffic to the storage account unless explicitly allowed via outbound rules or private endpoints. Azure AI Foundry hubs enforce network isolation for security, and if the storage account is not configured as an approved destination, the hub cannot establish the required connections for data access.

Answer

The location 'eastus' does not support AI Foundry hubs

Answer

The storage account name contains invalid characters

Answer

The key vault is not configured with a firewall rule

Question 15

Which TWO actions should you take to ensure compliance with data residency requirements when using Azure AI services across multiple regions?

Accepted Answer

Configure the AI service's data residency settings to restrict data to the region. Option B is correct because Azure AI services expose data residency/processing-location settings (for example, restricting processing to a specific geography or region) that let you keep data within the required jurisdiction, which is exactly what a data residency requirement demands. Option E is correct because deploying separate Azure AI service instances in each required region ensures that data is created, processed, and stored locally in that region rather than being routed to or replicated across other geographies, which is the most direct way to satisfy per-region residency mandates. Option A is not appropriate because a global resource with geo-replication intentionally spreads data across regions, which conflicts with residency restrictions. Option C is not appropriate because Azure Front Door is a global HTTP load-balancing and CDN service that routes traffic for latency, not a mechanism for enforcing where data is stored or processed. Option D is not appropriate because simply disabling data replication does not by itself guarantee that processing and storage remain confined to the required region, and it is not a supported residency control for Azure AI services.

Answer

Use a global Azure AI service resource and configure geo-replication

Answer

Enable Azure Front Door to route traffic to the nearest region

Answer

Disable data replication for the AI service