AI-102 Practice Test 26 — 15 Questions

Question 1

Which TWO actions can you take to reduce the cost of using Azure OpenAI Service for a chat application?

Accepted Answer

Set the max_tokens parameter to a lower value.. Reducing the max_tokens parameter directly limits the number of tokens generated per API call, which lowers the cost since Azure OpenAI Service bills per token (both input and output). By capping the response length, you avoid paying for unnecessarily long completions.

Answer

Increase the frequency penalty.

Answer

Enable content filtering.

Answer

Increase the max_tokens parameter to allow longer responses.

Question 2

You are implementing a knowledge mining solution using Azure AI Search. The data source is a large Azure Cosmos DB collection containing customer support tickets. Each ticket has fields: ticket_id, description, category, and resolution. You need to ensure that the search index can support fuzzy search and autocomplete suggestions. What should you configure in the index definition?

Accepted Answer

Set the 'searchable' attribute on the description field and define a suggester. Fuzzy search requires the 'searchable' attribute on fields to enable full-text search, and autocomplete suggestions require a 'suggester' configured on the index. The suggester defines which fields are used to generate suggestion candidates, and the 'searchable' attribute allows the description field to be tokenized and matched against partial or misspelled queries.

Answer

Set the 'filterable' attribute on the description field

Answer

Set the 'sortable' attribute on the ticket_id field

Answer

Set the 'facetable' attribute on the category field

Question 3

Which THREE practices should be followed to secure an Azure AI solution that uses Azure OpenAI Service and Azure AI Search?

Accepted Answer

Use managed identities to authenticate between Azure OpenAI and Azure AI Search.. Managed identities allow Azure OpenAI and Azure AI Search to authenticate securely without storing or rotating credentials. By assigning a managed identity to the Azure OpenAI service, you can grant it access to Azure AI Search via Azure RBAC, eliminating the need for API keys or shared access keys in code or configuration.

Answer

Store API keys in Azure Key Vault but use them directly in application code.

Answer

Place all AI services in a DMZ subnet with public IP addresses.

Question 4

You are the Azure AI engineer for a large e-commerce company. The company uses Azure Computer Vision to automatically tag product images uploaded by sellers. The system has been running smoothly for months. However, after a recent update to the Computer Vision API, you notice that certain images of clothing items are being tagged with incorrect labels, such as 'shoe' for a shirt. The images are clear and well-lit. You have confirmed that the image format (JPEG) is supported and the size is within limits. The issue occurs consistently for clothing items with similar colors. Other product categories work fine. You suspect the issue is related to the API version. What should you do first?

Accepted Answer

Check the API version used in the application code and compare with the latest version.. The issue began after a Computer Vision API update, and the problem is specific to certain clothing images with similar colors, indicating a potential regression or behavioral change in the API version. Checking the API version used in the application code against the latest version is the first logical troubleshooting step to identify if a breaking change or bug was introduced. This aligns with Azure AI best practices: always verify API version compatibility before modifying thresholds or retraining models.

Answer

Increase the image size limit.

Answer

Switch to a custom model trained on clothing items.

Answer

Reduce the confidence threshold to 50% to see if more tags appear.

Question 5

You are building a knowledge mining solution that uses Azure Cognitive Search and Azure AI Language. The solution must extract key phrases and detect the language of documents. Which THREE components are required?

Accepted Answer

A search index that contains fields for the extracted data.. A search index is the destination where extracted data (key phrases and language) must be stored for querying. Without an index, the extracted information has no structured location to be persisted and made searchable. The index schema must include fields specifically mapped to the output of the skillset's Key Phrase Extraction and Language Detection skills.

Answer

A custom skill to combine key phrases and language.

Answer

An indexer that runs on a schedule.

Question 6

You are building a multi-tenant application that uses Azure OpenAI. Each tenant has different content filtering requirements. How should you configure the solution to meet these requirements?

Accepted Answer

Create separate Azure OpenAI deployments for each tenant with their own content filter configurations.. Azure OpenAI content filters are configured at the deployment level, not at the application or API-call level. By creating separate deployments for each tenant, you can assign distinct content filter configurations (e.g., severity thresholds for hate, violence, self-harm) that are enforced server-side by Azure's content filtering service, ensuring isolation and compliance with each tenant's requirements.

Answer

Use a single Azure OpenAI deployment and configure content filters at the application level.

Answer

Configure different system messages for each tenant to enforce content policies.

Answer

Pass the tenant ID in the API call and use a custom middleware to apply filters.

Question 7

Your company is using Azure OpenAI Service to generate marketing copy. The compliance team requires that all generated content be reviewed for sensitive data before delivery. You need to implement a solution that automatically scans the output for personally identifiable information (PII) and blocks it if detected. Which service should you integrate?

Accepted Answer

Azure AI Content Safety with a custom blocklist. Azure AI Content Safety with a custom blocklist is the correct choice because it is specifically designed to filter and block harmful or sensitive content, including PII, in real-time. By adding a custom blocklist of PII patterns (e.g., regex for social security numbers), you can automatically block any output from Azure OpenAI that matches those patterns before delivery. This service integrates directly with Azure OpenAI Service via the content filtering system, making it the most appropriate for this use case.

Answer

Microsoft Purview Information Protection

Answer

Microsoft Defender for Cloud Apps

Answer

Azure AI Language PII detection

Question 8

Refer to the exhibit. You are deploying a GPT-4 model using Azure OpenAI Service. The deployment uses the Standard scale type. Which statement is true about this deployment?

Accepted Answer

The deployment uses pay-as-you-go pricing and global rate limits.. The Standard scale type in Azure OpenAI Service uses pay-as-you-go pricing, where you are billed based on the number of tokens processed. It also enforces global rate limits (e.g., tokens per minute) that apply across all deployments in the region, rather than providing dedicated capacity. Option D correctly identifies these characteristics.

Answer

The model version is not specified and will default to the latest.

Answer

Content filtering is disabled for this deployment.

Answer

The deployment uses provisioned throughput with reserved capacity.

Question 9

Refer to the exhibit. You are deploying a generative AI model as an online endpoint in Azure Machine Learning. You receive complaints that the endpoint returns 503 errors during peak hours. What is the most likely cause?

Accepted Answer

The manual scale setting with 2 instances may be insufficient for peak traffic.. 503 errors during peak hours indicate that the endpoint is overwhelmed by the request volume. With manual scaling set to only 2 instances, the compute capacity is insufficient to handle the increased traffic, causing the service to reject requests. Azure Machine Learning online endpoints require sufficient instance count or autoscaling to absorb traffic spikes.

Answer

The request timeout of 30 seconds is too short.

Answer

The environment variable MODEL_CACHE_SIZE is set too low.

Answer

The model version is not specified correctly.

Question 10

Your organization uses Azure AI Document Intelligence to extract data from invoices. The extraction accuracy for total amounts is low. You have a labeled dataset of 500 invoices. You need to improve the model's accuracy for the 'total amount' field. What should you do?

Accepted Answer

Create a custom neural model and train it with the labeled dataset.. Azure AI Document Intelligence's custom neural model is specifically designed to improve extraction accuracy for fields like 'total amount' by training on labeled datasets. Unlike the prebuilt invoice model, a custom neural model learns the unique layout and variations in your invoices, directly addressing low accuracy for a specific field. Training with 500 labeled invoices provides sufficient data to fine-tune the model's extraction capabilities.

Answer

Add additional predefined models for invoice processing.

Answer

Enable OCR enhancement to improve text recognition.

Answer

Increase the confidence threshold for the total amount field.

Question 11

Which TWO options are valid ways to reduce the cost of using Azure OpenAI Service?

Accepted Answer

Use provisioned throughput with reserved capacity.. Provisioned throughput with reserved capacity allows you to commit to a specific amount of throughput (tokens per minute) for a fixed period, typically one month or one year, in exchange for a significant discount compared to pay-as-you-go pricing. This reduces per-token costs when you have predictable workloads.

Answer

Increase the temperature parameter.

Answer

Increase the max_tokens parameter to get longer responses.

Answer

Enable content filtering on all requests.

Question 12

A company uses Azure AI Speech to provide real-time transcription for customer support calls. The solution must handle multiple languages and filter out profanity. Which THREE configurations are needed?

Accepted Answer

Set the SpeechConfig.SpeechRecognitionLanguage property. To achieve real-time transcription with multi-language support and profanity filtering using Azure AI Speech, three configurations are required. First, set the `SpeechConfig.SpeechRecognitionLanguage` property (Option C) to specify the source language for accurate transcription. Second, use the Speech SDK with intermediate results (Option D) to provide real-time feedback during the call. Third, enable the ProfanityFilter option in the Speech SDK (Option E) to filter out profanity. Option A (Batch Transcription) is incorrect because it is not real-time. Option B (custom speech model per language) is unnecessary because the standard prebuilt models support multiple languages and profanity filtering without customization.

Answer

Use the Batch Transcription REST API

Answer

Deploy a custom speech model for each language

Question 13

You are building an Azure AI Search solution that indexes data from multiple sources, including SQL Database and Azure Blob Storage. The index must be updated within 15 minutes of any source change. Which approach should you use to achieve near-real-time indexing?

Accepted Answer

Use the push API to send updates as soon as data changes. The push API (Azure Cognitive Search REST API or SDK) allows you to directly upload documents to the index as soon as data changes occur, bypassing the indexer's polling cycle. This provides sub-minute latency, meeting the 15-minute near-real-time requirement. Indexers with schedules or enrichment pipelines introduce inherent delays and are not designed for sub-minute updates.

Answer

Enable incremental enrichment on the skillset

Answer

Use an indexer with a schedule set to run every 5 minutes

Answer

Enable semantic search to speed up indexing

Question 14

Which TWO actions should you take to ensure that an Azure AI Search indexer can access data from an Azure Storage account that contains sensitive data?

Accepted Answer

Configure the search service to use a system-assigned managed identity. To ensure that an Azure AI Search indexer can access data from an Azure Storage account containing sensitive data, you should use a system-assigned managed identity for the search service (Option C) and allow the search service's IP address in the storage account firewall (Option D). Using a managed identity eliminates the need for keys and provides secure authentication. Allowing the search service's IP ensures that the indexer can connect through the firewall. Option A is incorrect because a private endpoint is for accessing the storage from a virtual network, not for indexer access. Option B is incorrect because shared access keys are less secure and expose the storage account. Option E is incorrect because disabling the firewall entirely is insecure.

Answer

Create a private endpoint in the storage account for the search service

Answer

Use a shared access key (SAS) in the data source definition

Answer

Disable the storage account firewall entirely

Question 15

Drag and drop the steps to deploy a custom language model using Azure AI Language into the correct order.

Accepted Answer

Step 1: Have labeled data ready, Step 2: Create the Azure AI Language resource, Step 3: Train the custom language model, Step 4: Evaluate the model performance, Step 5: Deploy the model to an endpoint. First, have labeled data ready, then create the resource, train the model, evaluate, and deploy.

Answer

Step 1: Create the Azure AI Language resource, Step 2: Have labeled data ready, Step 3: Train the custom language model, Step 4: Evaluate the model performance, Step 5: Deploy the model to an endpoint

Answer

Step 1: Have labeled data ready, Step 2: Create the Azure AI Language resource, Step 3: Evaluate the model performance, Step 4: Train the custom language model, Step 5: Deploy the model to an endpoint

Answer

Step 1: Have labeled data ready, Step 2: Create the Azure AI Language resource, Step 3: Train the custom language model, Step 4: Deploy the model to an endpoint, Step 5: Evaluate the model performance