Courseiva

Microsoft Azure AI Engineer Associate AI-102 (AI-102) — Questions 151–225

988 questions total · 14pages · All types, answers revealed

Take a mock examExam hub
Page 2

Page 3 of 14

Page 4
151
Multi-Selectmedium

Which THREE components are part of the Azure AI Bot Service?

Select 3 answers
A.Azure AI Search
B.Azure Bot Service (hosting)
C.Azure AI Language (CLU)
D.Bot Framework SDK
E.Bot Framework Composer
AnswersB, D, E

Azure Bot Service provides hosting for bots.

Why this answer

Azure Bot Service is the hosting component that provides a managed environment for deploying and running bots, integrating with channels like Teams, Slack, and Web Chat. It handles authentication, state management, and scaling, making it a core part of the Azure AI Bot Service.

Exam trap

The trap here is that candidates often confuse external AI services like Azure AI Search or Azure AI Language (CLU) as being part of the Azure AI Bot Service, when they are actually separate services that can be integrated but are not core components of the bot service itself.

Full explanation →
152
MCQhard

You are building an agentic solution that needs to perform actions on behalf of the user, such as sending emails and updating calendars. Which authentication approach should you use for the agent to access Microsoft Graph API with delegated permissions?

A.OAuth 2.0 client credentials flow
B.OAuth 2.0 authorization code flow with PKCE
C.OAuth 2.0 implicit flow
D.OAuth 2.0 device code flow
AnswerB

This flow is secure and allows delegated permissions.

Why this answer

OAuth 2.0 authorization code flow with PKCE is the recommended approach for delegated permissions in native apps that need to act on behalf of a user. Client credentials flow is for app-only access. Implicit flow is deprecated.

Device code flow is for devices without a browser.

Full explanation →
153
MCQhard

You are deploying a computer vision model using Azure AI Custom Vision with a small dataset of 200 images per class. The model shows high accuracy on training data but low accuracy on test data. Which action should you take to reduce overfitting?

A.Increase the learning rate
B.Reduce the image size to lower resolution
C.Increase the number of training epochs
D.Increase the training dataset size with more varied images
AnswerD

More data helps the model generalize better and reduces overfitting.

Why this answer

The correct answer is to increase the training dataset size by adding more varied images. Overfitting occurs when the model learns noise from a small dataset. Option A is wrong because increasing the number of training epochs can actually increase overfitting.

Option B is wrong because increasing the learning rate may cause divergence. Option D is wrong because reducing image size can lose important features.

Full explanation →
154
MCQeasy

Your application needs to determine whether two photos of the same person are of the same individual, even if they are from different angles. Which Azure AI service should you use?

A.Azure AI Video Indexer
B.Azure AI Custom Vision
C.Azure AI Vision OCR
D.Azure AI Face
AnswerD

Face verification is designed for this purpose.

Why this answer

Azure AI Face's 'verify' operation compares two faces and returns a confidence score for whether they belong to the same person.

Full explanation →
155
MCQmedium

You are using Azure AI Document Intelligence to process a large batch of PDF forms. The forms have varying layouts and handwriting. You need to extract text and key-value pairs. Which custom model type should you train?

A.Custom template model
B.Prebuilt-layout model
C.Custom neural model
D.Custom composed model
AnswerC

Neural models handle varying layouts and handwriting better.

Why this answer

Option B is correct because a template model works best with forms that have a fixed layout, but since layouts vary, a neural model is more robust. Option A is wrong because a template model assumes fixed layout. Option C is wrong because a composed model combines multiple models but is not the primary choice for varying layouts.

Option D is wrong because a prebuilt model is not customized for your forms.

Full explanation →
156
Multi-Selectmedium

You are designing an agentic solution using Azure AI Agent Service. The agent needs to be able to both read and write data to an Azure SQL database. Which TWO tools should you configure?

Select 2 answers
A.Function calling
B.KQL
C.Grounding with Bing
D.Code Interpreter
E.Knowledge base
AnswersA, D

Function calling can be used to define custom functions that interact with the database.

Why this answer

Code Interpreter can execute SQL queries via Python. Function calling can be used to invoke custom database operations. Knowledge base is for search, not write.

KQL is for Azure Data Explorer.

Full explanation →
157
MCQmedium

A company is developing an agent that uses Azure AI Language to extract entities and intents from user queries. The agent receives a query: 'Book a flight to Paris on Friday.' The agent should extract the intent as 'BookFlight' and entities as 'Paris' (destination) and 'Friday' (date). The team uses a custom entity extraction model. After testing, the model extracts 'Paris' as location but fails to extract 'Friday' as date. What should the team do to fix this?

A.Increase the training data for location entities.
B.Add a prebuilt entity component for date.
C.Train the entity extraction model with more examples of dates.
D.Use a different intent classification model.
AnswerC

More training examples improve entity recognition.

Why this answer

Option C is correct because the custom entity extraction model in Azure AI Language requires sufficient labeled examples for each custom entity type to learn patterns. Since the model extracts 'Paris' (location) but fails on 'Friday' (date), the issue is specifically with the date entity's training data, not the location. Adding more diverse examples of date expressions (e.g., 'next Monday', 'tomorrow', 'March 5th') will improve the model's ability to recognize 'Friday' as a date entity.

Exam trap

The trap here is that candidates confuse the need for more training data (Option C) with the use of prebuilt components (Option B), assuming prebuilt entities can fix custom model gaps, but prebuilt entities are not part of the custom entity extraction pipeline and would require a different project type.

How to eliminate wrong answers

Option A is wrong because increasing training data for location entities does not address the failure to extract date entities; the location entity already works correctly. Option B is wrong because adding a prebuilt entity component for date would override the custom entity extraction model's behavior, which contradicts the requirement to use a custom entity extraction model; prebuilt entities are used in orchestration workflows, not to fix custom model training deficiencies. Option D is wrong because the intent classification model ('BookFlight') is correctly identifying the intent; the problem lies with entity extraction, not intent classification, so changing the intent model would not resolve the date extraction failure.

Full explanation →
158
MCQeasy

You are building a chatbot using Azure AI Bot Service and Language Service. The bot must recognize user intent for 'check order status'. How should you configure the Language Service?

A.Create a custom intent classification project
B.Deploy a QnA Maker knowledge base
C.Configure a sentiment analysis endpoint
D.Use the prebuilt entity extraction model
AnswerA

Intent classification trains a model to map utterances to intents.

Why this answer

To recognize user intent for 'check order status', you need a custom intent classification project in Azure Language Service. This project type uses a trained model to map utterances to specific intents, such as 'CheckOrderStatus', which is exactly what the chatbot requires. Prebuilt models or QnA Maker do not provide custom intent recognition.

Exam trap

The trap here is that candidates confuse intent recognition with entity extraction or QnA, assuming any Language Service feature can handle intents, but only custom intent classification (or conversational language understanding) is designed for this purpose.

How to eliminate wrong answers

Option B is wrong because QnA Maker is designed for question-answering over a knowledge base, not for classifying user intent from natural language utterances. Option C is wrong because sentiment analysis determines the emotional tone of text, not the user's intent or goal. Option D is wrong because prebuilt entity extraction identifies named entities like dates or locations but does not classify the overall intent of a user message.

Full explanation →
159
MCQmedium

You are deploying an Azure AI Language service solution for a multilingual customer support chatbot. The solution must support real-time translation between English, Spanish, and French. Which Azure resource should you provision?

A.Azure AI Speech
B.Azure OpenAI Service
C.Azure AI Language
D.Azure AI Translator
AnswerD

Translator resource is designed for real-time text translation.

Why this answer

Azure AI Translator is the correct resource because it provides real-time text translation across multiple languages, including English, Spanish, and French, via a dedicated REST API. The scenario specifically requires translation between languages, not speech recognition or generative AI, making Azure AI Translator the precise service for this task.

Exam trap

The trap here is that candidates confuse Azure AI Language with Azure AI Translator, assuming the 'Language' service includes translation, when in fact translation is a separate service under the Azure AI Services umbrella.

How to eliminate wrong answers

Option A is wrong because Azure AI Speech handles speech-to-text, text-to-speech, and speech translation, but it is not optimized for pure text translation between multiple languages without audio input. Option B is wrong because Azure OpenAI Service is designed for generative AI tasks like content creation and conversation, not for direct, real-time text translation between specific languages. Option C is wrong because Azure AI Language provides natural language processing capabilities such as sentiment analysis and entity recognition, but it does not include a dedicated real-time translation API; translation is handled by Azure AI Translator.

Full explanation →
160
MCQeasy

You are using Azure OpenAI Service to generate marketing copy. You notice that the output sometimes contains factual inaccuracies about your company's products. Which action can you take to improve factual accuracy?

A.Lower the temperature to 0.
B.Include relevant product information in the system message.
C.Increase the maxTokens to 4000.
D.Add a stop sequence to limit output.
AnswerB

Providing accurate context in the prompt helps the model generate factual responses.

Why this answer

Option A is correct because providing ground truth data in the prompt (e.g., via RAG) improves accuracy. Option B is wrong as temperature affects creativity. Option C is wrong as maxTokens limits length.

Option D is wrong as stop sequences control generation stopping.

Full explanation →
161
Multi-Selectmedium

Which TWO Azure AI services can be used to detect objects in images?

Select 2 answers
A.Video Indexer
B.Face API
C.Custom Vision
D.Azure AI Document Intelligence
E.Computer Vision Object Detection API
AnswersC, E

Custom Vision can be trained for object detection.

Why this answer

Option A and B are correct because Custom Vision and Computer Vision Object Detection both support object detection. Option C is wrong because Face API detects faces, not generic objects. Option D is wrong because Document Intelligence extracts text and structure.

Option E is wrong because Video Indexer analyzes video, not images.

Full explanation →
162
MCQeasy

Your company uses Azure AI Search to power a customer support portal. The search index includes product documentation and known issues. Recently, the portal's search performance has degraded, and users report slow response times. You need to identify the cause of the performance issue. What should you check first?

A.Review the search service metrics for high query latency and CPU usage.
B.Check the size of the index storage in the Azure portal.
C.Ensure the index schema does not have too many fields.
D.Verify that the skillset is not running during peak hours.
AnswerA

Metrics help identify performance bottlenecks.

Why this answer

High query latency in Azure AI Search is often caused by high CPU utilization in the search service. Checking metrics in the Azure portal can reveal if the service is overburdened. Option A is wrong because storage size does not directly cause slow queries.

Option C is wrong because the skillset is not involved unless enrichment is ongoing. Option D is wrong because the index schema size is not a primary performance factor.

Full explanation →
163
MCQeasy

Your organization is deploying Azure AI Document Intelligence to process invoices. You need to ensure that the solution meets compliance requirements by preventing data from being stored outside the European Union. What should you configure?

A.Deploy Azure AI Document Intelligence in a multi-region configuration across EU datacenters.
B.Create the Azure AI Document Intelligence resource in a region within the European Union and configure data residency settings.
C.Enable the Resource Firewall on the Azure AI Document Intelligence resource.
D.Use a Private Endpoint to connect to the Azure AI Document Intelligence resource.
AnswerB

Creating the resource in the EU and configuring data residency ensures data stays within the EU.

Why this answer

Option B is correct because creating the Azure AI Document Intelligence resource in an EU region and configuring data residency settings ensures that all data processed by the service remains within the European Union, meeting compliance requirements. Azure AI Document Intelligence does not automatically guarantee data residency based solely on the resource region; explicit data residency configuration is required to prevent data from being stored or processed outside the specified geographic boundary.

Exam trap

The trap here is that candidates assume simply creating the resource in an EU region automatically ensures data residency, but Azure AI Document Intelligence requires explicit data residency configuration to prevent data from being stored or processed outside the chosen region.

How to eliminate wrong answers

Option A is wrong because deploying in a multi-region configuration across EU datacenters does not prevent data from being stored outside the EU; it may replicate data across regions for redundancy, potentially violating data residency requirements. Option C is wrong because enabling the Resource Firewall restricts network access to the resource but does not control where data is stored or processed geographically. Option D is wrong because using a Private Endpoint ensures private network connectivity to the resource but does not enforce data residency; data can still be processed or stored in any region where the resource is deployed.

Full explanation →
164
MCQmedium

You are calling the Azure AI Vision Image Analysis API with the above JSON request. The API returns a response that includes tags, objects, and a caption. However, the caption does not describe the image accurately. Which change should you make to improve the caption quality?

A.Change model-version to a newer version
B.Add 'read' feature to the request
C.Set gender-neutral-caption to false
D.Set language to 'en-US'
AnswerC

Gender-neutral captions may omit descriptive terms, reducing accuracy.

Why this answer

The correct answer is to remove 'gender-neutral-caption' or set it to false. Gender-neutral captions may be less descriptive. Option A is wrong because the model-version is already latest.

Option B is wrong because language is already English. Option D is wrong because adding 'read' feature is for OCR, not caption quality.

Full explanation →
165
Multi-Selecthard

Which THREE factors should you consider when choosing between Azure OpenAI Service and Azure Machine Learning for deploying a generative AI model?

Select 3 answers
A.Integration with Microsoft Purview for data governance.
B.Latency requirements: Azure OpenAI may offer lower latency for standard models.
C.Ability to scale to thousands of concurrent requests.
D.Need for custom model architecture: Azure ML supports custom models, Azure OpenAI uses pre-trained.
E.Operational overhead: Azure OpenAI is a fully managed service.
AnswersB, D, E

Azure OpenAI endpoints are optimized for low latency, whereas Azure ML may require additional optimization.

Why this answer

Options A, C, and E are correct. A: Azure OpenAI is fully managed with less operational overhead. C: Azure OpenAI offers pre-trained models, while Azure ML allows custom models.

E: Azure OpenAI may have lower latency for similar model sizes. B is wrong because both services support scaling. D is wrong because both services support governance.

Full explanation →
166
MCQmedium

You are building a solution to analyze customer feedback from multiple sources: emails, chat logs, and survey responses. You need to detect the overall sentiment trend over time and identify the most frequently mentioned topics. The solution must also allow the business analyst to ask natural language questions about the data (e.g., 'Show me complaints about shipping in the last month'). You have all data in Azure Blob Storage. You need to implement a solution with minimal custom code. Which combination of Azure services should you use?

A.Use Azure OpenAI Service to analyze sentiment and generate summaries, and store results in a Cosmos DB for querying.
B.Use Azure AI Language to extract sentiment and key phrases, then index the data in Azure Cognitive Search with semantic search enabled; use the search's built-in features for trend analysis and natural language queries.
C.Use Azure AI Language to perform sentiment analysis and key phrase extraction, then load the results into Power BI for trend analysis and natural language Q&A.
D.Use Azure AI Language's custom question answering to create a knowledge base from the feedback and allow natural language queries.
AnswerB

Azure Cognitive Search can index the feedback with extracted metadata, and its semantic search can interpret natural language queries like 'complaints about shipping' and return relevant documents.

Why this answer

Option C is correct because Azure AI Language can perform sentiment analysis and key phrase extraction, and Azure Cognitive Search with semantic ranker can enable natural language querying over the indexed data. Option A is wrong because Azure OpenAI Service alone does not provide built-in analytics over stored data. Option B is wrong because Power BI is for visualization, not natural language querying over text.

Option D is wrong because Azure AI Language's question answering is for FAQ, not for querying indexed data.

Full explanation →
167
MCQeasy

A company uses Azure Custom Vision to classify images of defective parts. After deploying the model, the accuracy is low. The team only has 10 images per class. What is the most effective way to improve accuracy?

A.Use a different classification algorithm.
B.Add at least 50 more images per class with variations.
C.Reduce the image resolution to speed up training.
D.Increase the number of training iterations (epochs).
AnswerB

More data improves model accuracy.

Why this answer

Option B is correct because Azure Custom Vision relies on deep learning models that require a sufficient number of diverse training images to generalize well. With only 10 images per class, the model is severely underfit and prone to overfitting; adding at least 50 more images per class with variations in lighting, angle, and background provides the necessary data diversity to improve accuracy significantly.

Exam trap

The trap here is that candidates often assume increasing epochs or changing the algorithm will fix low accuracy, but the real bottleneck is insufficient and non-diverse training data, which is the most common cause of poor Custom Vision model performance.

How to eliminate wrong answers

Option A is wrong because Azure Custom Vision automatically selects and tunes the underlying classification algorithm (a convolutional neural network) based on the dataset; manually changing the algorithm is not supported and would not address the core issue of insufficient training data. Option C is wrong because reducing image resolution can discard important fine-grained features needed to detect defects, and Azure Custom Vision already resizes images to a fixed input size (e.g., 224x224) during training, so further reduction harms accuracy rather than improving it. Option D is wrong because increasing training iterations (epochs) with only 10 images per class will cause the model to overfit to the small dataset, memorizing the training examples rather than learning generalizable patterns, leading to poor accuracy on new images.

Full explanation →
168
Multi-Selecthard

Which THREE components are required to build a knowledge mining solution using Azure AI Search that extracts and enriches content from PDF files?

Select 4 answers
A.Index
B.Skillset
C.Semantic configuration
D.Data source
E.Indexer
AnswersA, B, D, E

Stores the enriched data.

Why this answer

Option A is correct because a data source defines where to pull data. Option C is correct because an index stores the enriched content. Option E is correct because a skillset defines the enrichment pipeline.

Option B is wrong because an indexer is required, but a datasource and index are separate components; an indexer connects them. Option D is wrong because a semantic configuration is optional.

Full explanation →
169
MCQhard

You are designing an NLP solution to analyze legal documents. The solution must identify specific clauses and parties involved. Which Azure AI service is most appropriate?

A.Custom Named Entity Recognition in Azure AI Language
B.Pre-built Named Entity Recognition in Azure AI Language
C.Text Analytics for Health
D.Immersive Reader
AnswerA

Custom NER can be trained to recognize domain-specific entities.

Why this answer

Option B is correct because Custom NER allows you to train a model to recognize custom entities like clauses and parties. Option A is incorrect because pre-built NER only recognizes common entities. Option C is incorrect because text analytics for health is focused on medical data.

Option D is incorrect because the Immersive Reader is for reading support, not entity extraction.

Full explanation →
170
Multi-Selecteasy

Which TWO features of Azure AI Search allow you to improve the relevance of search results for users?

Select 2 answers
A.Synonym maps
B.Semantic search
C.Suggesters
D.Scoring profiles
E.Filterable fields
AnswersB, D

Semantic search uses AI to understand the intent of the query and improve ranking.

Why this answer

Options B and D are correct. B: Scoring profiles allow you to boost results based on criteria. D: Semantic search improves relevance by understanding query intent.

Option A is wrong because filterable does not affect ranking. Option C is wrong because suggesters are for autocomplete, not relevance. Option E is wrong because synonym maps expand queries but do not directly improve relevance ranking.

Full explanation →
171
MCQhard

You are deploying a custom text classification model using Azure AI Language. The model must be retrained monthly with new labeled data. You need to automate the retraining process with minimal manual intervention. Which approach should you use?

A.Create an Azure DevOps pipeline that manually retrains the model every month
B.Retrain the model manually using Language Studio each month
C.Use the Azure AI Language REST API to trigger training and deployment on a schedule using Azure Logic Apps
D.Use Azure Machine Learning to host the custom model and automate retraining
AnswerC

Logic Apps can call the Language API on a schedule to automate retraining.

Why this answer

Option C is correct because Azure Logic Apps can schedule HTTP requests to the Azure AI Language REST API to trigger training and deployment of a custom text classification model. This approach automates the monthly retraining process without manual intervention, using the API's capabilities for model creation, training, and deployment.

Exam trap

The trap here is that candidates may confuse Azure Machine Learning with Azure AI Language, thinking that Azure Machine Learning is the appropriate service for hosting and retraining custom text classification models, when in fact Azure AI Language provides its own REST API for this purpose and is the correct service for custom text classification.

How to eliminate wrong answers

Option A is wrong because an Azure DevOps pipeline that manually retrains the model every month still requires manual intervention to trigger the pipeline, contradicting the requirement for minimal manual intervention. Option B is wrong because retraining manually using Language Studio each month is entirely manual and does not automate the process. Option D is wrong because Azure Machine Learning is not designed to host custom text classification models built with Azure AI Language; it is a separate platform for building and deploying machine learning models, and using it would introduce unnecessary complexity and integration overhead.

Full explanation →
172
MCQeasy

You are designing a solution that uses Azure AI Language to analyze customer feedback. The solution must detect sentiment, extract key phrases, and identify named entities. Which feature should you use?

A.Azure AI Language service
B.Azure AI Speech service
C.Azure AI Computer Vision
D.Translator API
AnswerA

Azure AI Language provides sentiment, key phrases, and entity recognition.

Why this answer

The Azure AI Language service provides pre-built capabilities for sentiment analysis, key phrase extraction, and named entity recognition (NER) as part of its text analytics features. These three tasks are directly supported by the service's Analyze API, making it the correct choice for analyzing customer feedback text.

Exam trap

The trap here is that candidates may confuse Azure AI Language with other Azure AI services that have overlapping names (e.g., Translator API for language tasks) or assume Speech or Vision services can perform text analysis, but only the Language service provides the specific trio of sentiment, key phrases, and NER.

How to eliminate wrong answers

Option B is wrong because Azure AI Speech service is designed for speech-to-text, text-to-speech, and speech translation, not for analyzing text sentiment, key phrases, or named entities. Option C is wrong because Azure AI Computer Vision focuses on image and video analysis (e.g., object detection, OCR), not on natural language processing tasks like sentiment or entity extraction. Option D is wrong because Translator API is specifically for machine translation between languages and does not include sentiment analysis, key phrase extraction, or named entity recognition.

Full explanation →
173
MCQhard

You are using Azure AI Translator to translate documents from English to French. Some technical terms must remain untranslated. How should you handle this?

A.Set the includeUntranslated parameter to true
B.Train a custom translation model that ignores those terms
C.Post-process the output to revert translations of those terms
D.Provide a dictionary with the terms and their translations set to the same word
AnswerD

The dictionary allows forcing a specific translation; setting the target same as source prevents translation.

Why this answer

Azure AI Translator supports a dictionary to specify words that should not be translated. The 'includeUntranslated' parameter does not exist. Post-editing is inefficient.

Using a custom model requires training data.

Full explanation →
174
MCQhard

You are implementing a solution that uses Azure AI Vision to analyze images in a batch processing pipeline. The pipeline processes thousands of images daily. You need to minimize cost while ensuring that all images are processed within 24 hours. Which service tier should you choose?

A.Free tier F0
B.Standard S1 tier
C.Standard S0 tier
D.Standard S2 tier
AnswerB

S1 provides higher throughput and pay-per-transaction cost, ideal for batch processing.

Why this answer

The Standard S1 tier provides pay-per-transaction pricing, which is cost-effective for high-volume batch processing. Free tier has limits, and S0 is too slow.

Full explanation →
175
Multi-Selecteasy

A company is building a knowledge mining solution using Azure AI Search. They need to extract entities from a large set of PDF documents stored in Azure Blob Storage. The solution must use a built-in AI skill to identify people, organizations, and locations. Which TWO actions should be taken? (Choose two.)

Select 2 answers
A.Configure the index to use a custom analyzer.
B.Add the Key Phrase Extraction skill to the skillset.
C.Deploy a custom skill using Azure Functions to extract entities.
D.Enable OCR (Optical Character Recognition) in the indexer configuration.
E.Add the Entity Recognition skill to the skillset.
AnswersD, E

OCR extracts text from PDFs so that the Entity Recognition skill can process it.

Why this answer

Option A is correct because the Entity Recognition skill identifies people, organizations, and locations. Option D is correct because the built-in OCR skill is needed to extract text from PDFs before entity recognition can occur. Options B, C, and E are either unrelated or unnecessary.

Full explanation →
176
MCQhard

You are a Microsoft AI engineer for a multinational retail company. The company uses Microsoft Copilot Studio to build an agent for employee self-service. The agent must handle three main tasks: (1) look up employee information from an on-premises HR database, (2) submit expense reports, and (3) answer questions about company policies stored in SharePoint Online. The HR database can only be accessed via a REST API that requires NTLM authentication. The expense report submission must be routed to a third-party system that uses OAuth 2.0. The policy documents are in multiple languages and the agent must provide answers in the user's language. The agent is published to Microsoft Teams and must support single sign-on (SSO) for authenticated users. The company has strict security requirements: all backend calls must use the user's identity, not a generic service account. The agent must also log all interactions for audit purposes. You need to design the solution architecture. Which combination of Azure services and configurations should you use?

A.Use Azure Functions to wrap the on-premises API and the expense system, connect via hybrid connections, and use the Bot Framework SDK to handle multi-language.
B.Use Power Automate flows for all backend calls, configure SharePoint as a knowledge source, and enable SSO with Microsoft Entra ID.
C.Use Azure Logic Apps with on-premises data gateway for the HR API, custom connector for expense system, and enable generative answers with SharePoint.
D.Use an on-premises data gateway with a custom connector for the HR API, a custom connector for the expense system with OAuth 2.0, enable generative answers with SharePoint, and configure SSO with Microsoft Entra ID.
AnswerD

This architecture meets all requirements: on-premises gateway for NTLM, custom connectors for authentication, generative answers for multi-language, and SSO for Teams.

Why this answer

Option D is correct because it addresses all requirements: on-premises data gateway for NTLM-secured API, custom connector for OAuth 2.0, generative answers with SharePoint for multi-language policies, and SSO with Microsoft Entra ID. Option A is wrong because Power Automate flows cannot directly use NTLM without a gateway. Option B is wrong because Azure Functions cannot directly access on-premises without a gateway.

Option C is wrong because Logic Apps need a gateway and custom connectors are better for OAuth 2.0.

Full explanation →
177
MCQmedium

You are building an agentic solution using Microsoft Semantic Kernel. The agent needs to orchestrate multiple plugins. One plugin returns a large dataset that exceeds the model's context window. What is the best way to handle this?

A.Split the data into multiple smaller API calls and combine results
B.Truncate the data to fit the context window
C.Configure the plugin to return only a subset of the data and mark the rest as sensitive
D.Use a summarization plugin to condense the data before passing it to the model
AnswerD

Summarization preserves key info and reduces size.

Why this answer

Option B is correct because summarizing the data reduces its size while preserving key information. Option A is incorrect because truncation may lose critical data. Option C is incorrect because splitting into multiple calls increases latency and tokens.

Option D is incorrect because marking as sensitive does not solve the size issue.

Full explanation →
178
MCQhard

Refer to the exhibit. You have created an assistant with the above configuration. When you send a message 'What is the weather in Seattle?', the assistant responds without calling the function. What is the most likely cause?

A.The 'instructions' are not being followed
B.The 'strict' parameter is set to true
C.The function is missing the 'description' property
D.The 'tool_resources' code_interpreter has empty file_ids
AnswerC

The description helps the model decide to call the function.

Why this answer

The function definition is missing the 'description' field, which is required for the model to understand when to call the function. The instructions say to call the function, but the model relies on the function's description. The other settings are correct.

Full explanation →
179
Multi-Selectmedium

Which TWO actions are required to enable a custom chatbot built with Azure OpenAI to answer questions based on a company's internal PDF documents?

Select 2 answers
A.Use Azure AI Document Intelligence to extract text from PDFs before indexing
B.Deploy Azure AI Content Safety to filter responses
C.Fine-tune the GPT model on the PDF content
D.Ingest the PDFs into an Azure Cognitive Search index
E.Configure the Azure OpenAI deployment to use 'Add your data' with the search index
AnswersD, E

Indexing enables retrieval of relevant content from PDFs.

Why this answer

Option D is correct because Azure Cognitive Search provides the indexing and retrieval capabilities needed to make PDF content searchable. By ingesting PDFs into an Azure Cognitive Search index, the chatbot can perform vector or keyword searches over the extracted text, enabling it to retrieve relevant passages to answer user questions. This is the standard approach for grounding a custom chatbot on proprietary documents without modifying the underlying model.

Exam trap

The trap here is that candidates often confuse fine-tuning (option C) with the RAG pattern, mistakenly believing they must retrain the model on proprietary data, when in fact the 'Add your data' feature with a search index is the correct and simpler approach for question-answering over internal documents.

Full explanation →
180
Multi-Selecteasy

You are building a solution to extract custom entities from legal contracts using Azure AI Language. You have a small set of labeled documents. Which two features should you use to build and improve the custom NER model? (Choose two.)

Select 2 answers
A.Use active learning to automatically suggest new labels from unlabeled documents.
B.Use the prebuilt NER model as a base and extend it with custom entities.
C.Configure an orchestration workflow to route documents to the best model.
D.Add synonyms for each entity to improve recognition of variations.
E.Extract key phrases from the documents and use them as features.
AnswersA, D

Active learning identifies uncertain predictions and suggests them for labeling, reducing manual effort.

Why this answer

Options A and B are correct because active learning helps label additional data efficiently, and using synonyms improves entity recognition for varied phrasings. Option C is wrong because key phrase extraction is not part of custom NER training. Option D is wrong because prebuilt NER cannot be extended with custom entities.

Option E is wrong because orchestration workflow is for intent detection, not NER.

Full explanation →
181
Multi-Selectmedium

Which TWO are valid ways to manage cost when using Azure OpenAI Service in a production application?

Select 2 answers
A.Fine-tune the model to reduce the number of examples needed in prompts
B.Increase the temperature parameter to 1.0
C.Use a smaller model like GPT-3.5-turbo instead of GPT-4 for simpler tasks
D.Provision more PTUs to get a lower rate per token
E.Set the max_tokens parameter to the minimum needed for the response
AnswersC, E

Smaller models have lower per-token costs.

Why this answer

Option C is correct because using a smaller model like GPT-3.5-turbo for simpler tasks directly reduces the per-token cost compared to GPT-4, which is significantly more expensive. Azure OpenAI Service charges based on model tier and token usage, so selecting the appropriate model for the task complexity is a primary cost management strategy.

Exam trap

The trap here is that candidates may confuse fine-tuning with prompt optimization, or assume that increasing PTUs lowers per-token cost, when in fact PTUs are a fixed-cost commitment that increases total expenditure.

Full explanation →
182
MCQeasy

A company uses Azure Content Moderator to moderate text in a chat application. They want to automatically reject messages that contain profanity or personal data. Which API should they use?

A.Review API
B.Video Moderation API
C.Image Moderation API
D.Text Moderation API
AnswerD

The Text Moderation API screens text for profanity and personally identifiable information (PII).

Why this answer

The Text Moderation API (D) is the correct choice because it is specifically designed to scan text content for profanity, personally identifiable information (PII), and other unwanted text patterns. This API returns a moderation score and a list of detected terms, enabling automated rejection of messages that violate the defined policies.

Exam trap

The trap here is that candidates may confuse the Review API with the moderation APIs, not realizing that the Review API is for manual review workflows rather than automated content detection.

How to eliminate wrong answers

Option A is wrong because the Review API is used for human-in-the-loop review workflows, not for automated detection and rejection of profanity or personal data. Option B is wrong because the Video Moderation API is designed to moderate video content, not text messages. Option C is wrong because the Image Moderation API handles image content, not text-based chat messages.

Full explanation →
183
MCQeasy

Your organization wants to implement a document processing pipeline that extracts text from scanned PDFs and identifies named entities. Which two Azure AI services should you use?

A.Azure AI Language
B.Azure AI Custom Vision
C.Azure AI Document Intelligence
D.Azure AI Translator
E.Azure AI Speech
AnswerA, C

Provides named entity recognition to identify entities in text.

Why this answer

Azure AI Document Intelligence (formerly Form Recognizer) is used to extract text from scanned PDFs via OCR, while Azure AI Language provides pre-built named entity recognition (NER) to identify entities like people, organizations, and locations. Together, they form a complete pipeline: Document Intelligence handles the image-to-text conversion, and Language processes the extracted text for entity extraction.

Exam trap

The trap here is that candidates often confuse Azure AI Document Intelligence with Azure AI Custom Vision, thinking Custom Vision can perform OCR, but Document Intelligence is the dedicated service for document text extraction and layout analysis.

How to eliminate wrong answers

Option B is wrong because Azure AI Custom Vision is designed for image classification and object detection, not for OCR or text extraction from scanned documents. Option D is wrong because Azure AI Translator is a machine translation service that converts text between languages, not for extracting text from images or identifying named entities. Option E is wrong because Azure AI Speech handles speech-to-text and text-to-speech, not OCR or NER from scanned PDFs.

Full explanation →
184
MCQhard

Refer to the exhibit. You have created a Text Analytics resource and retrieved its keys. You want to use the key1 to call the Sentiment Analysis API from a Python application. Which endpoint URL should you use?

A.https://mytextanalytics.cognitiveservices.azure.com/sentiment/v3.1
B.https://mytextanalytics.api.cognitive.microsoft.com/text/analytics/v3.1/sentiment
C.https://mytextanalytics.cognitiveservices.azure.com/analyze
D.https://mytextanalytics.cognitiveservices.azure.com/text/analytics/v3.1/sentiment
AnswerD

This is the correct endpoint for sentiment analysis.

Why this answer

The exhibit shows the endpoint is https://mytextanalytics.cognitiveservices.azure.com/. For Text Analytics, the sentiment endpoint is appended with /text/analytics/v3.1/sentiment. The correct URL is https://mytextanalytics.cognitiveservices.azure.com/text/analytics/v3.1/sentiment.

Full explanation →
185
MCQmedium

Your organization has a large corpus of legal documents stored in Azure Blob Storage. You need to build a solution that allows lawyers to ask natural language questions and get answers directly from the documents, without moving data out of Azure. Which service should you use?

A.Azure AI Document Intelligence
B.Azure AI Search with semantic search
C.Azure AI Language Service with custom question answering
D.Azure AI Computer Vision
AnswerB

Semantic search in Azure AI Search can understand natural language queries and return relevant passages.

Why this answer

Option C is correct because Azure AI Search with semantic search can index documents and return answers using natural language queries. Option A is wrong because Document Intelligence extracts structured data but does not provide Q&A. Option B is wrong because Language Service provides Q&A but requires a knowledge base.

Option D is wrong because Computer Vision is for image analysis.

Full explanation →
186
MCQeasy

A company wants to build a customer support agent using Microsoft Copilot Studio. The agent needs to understand natural language and handle complex queries beyond simple keyword matching. The agent should be able to escalate to a human agent when it cannot resolve the issue. Which feature should the agent use to understand natural language?

A.Create a Power Automate flow to process queries.
B.Enable generative answers and configure a knowledge source.
C.Create topics with trigger phrases.
D.Use entities to extract key information.
AnswerB

Generative answers use AI to understand natural language and provide responses from knowledge sources.

Why this answer

Option C is correct because generative answers use large language models to understand natural language and handle complex queries. Option A is wrong because topics are for structured conversations. Option B is wrong because entities extract specific data but not understand intent.

Option D is wrong because Power Automate flows are for backend actions, not understanding.

Full explanation →
187
MCQmedium

Refer to the exhibit. You are deploying an Azure AI Services resource using an ARM template. After deployment, you cannot access the resource from any client, including the Azure portal. What is the most likely cause?

A.The SKU S0 does not support network restrictions.
B.The cognitiveServiceName conflicts with an existing resource.
C.The networkAcls defaultAction is set to Deny with no allowed IP or virtual network rules.
D.The virtualNetworkRules array is empty.
AnswerC

Blocks all traffic, including portal.

Why this answer

The networkAcls defaultAction set to Deny with no allowed IP or virtual network rules blocks all traffic to the Azure AI Services resource, including requests from the Azure portal and any client. This is because the default network access control list (ACL) denies all traffic unless explicitly permitted by IP rules or virtual network rules, rendering the resource inaccessible even for management operations.

Exam trap

The trap here is that candidates often assume an empty virtualNetworkRules array means no restrictions, but the defaultAction property controls the default behavior, and setting it to Deny with no allow rules blocks all traffic regardless of empty arrays.

How to eliminate wrong answers

Option A is wrong because the S0 SKU does support network restrictions; network ACLs are available for all paid SKUs (S0 and above), and the issue is not related to SKU limitations. Option B is wrong because a name conflict would cause a deployment error, not a scenario where the resource is deployed but inaccessible; the ARM template would fail with a conflict error during deployment. Option D is wrong because an empty virtualNetworkRules array is irrelevant when the defaultAction is Deny; the resource would still be blocked since no IP rules are specified to allow traffic, and an empty array does not implicitly permit any traffic.

Full explanation →
188
MCQeasy

You are using Azure AI Language to analyze customer reviews. You need to determine whether each review expresses a positive, negative, or neutral sentiment. Which API should you call?

A.Language detection API.
B.Entity recognition API.
C.Sentiment analysis API.
D.Key phrase extraction API.
AnswerC

Returns sentiment labels and scores.

Why this answer

Option B is correct because the sentiment analysis API in Azure AI Language returns sentiment labels (positive, negative, neutral) and confidence scores. Option A is wrong because the language detection API detects language. Option C is wrong because the entity recognition API extracts entities.

Option D is wrong because the key phrase extraction API extracts key phrases.

Full explanation →
189
MCQmedium

You are testing an Azure OpenAI Service chat completion with function calling. The assistant returned null content and a tool call. What does this indicate?

A.The model is unable to answer the question and returned an error
B.The system message is preventing the model from responding
C.The model is using a tool to answer the question, so content is null
D.The function call syntax is invalid and the model skipped it
AnswerC

Null content with tool call indicates the model is delegating to a function.

Why this answer

Option C is correct because in Azure OpenAI Service function calling, when the model determines that it needs to invoke a tool (e.g., an external API or database query) to fulfill the user's request, it returns a `tool_calls` object with `content` set to `null`. This is the expected behavior—the model is signaling that it is delegating the response to the tool, not that it has failed or is blocked.

Exam trap

The trap here is that candidates often misinterpret a null content as a model failure or error, when in fact it is a standard indicator that the model is actively using a tool to fulfill the request.

How to eliminate wrong answers

Option A is wrong because a null content with a tool call is not an error; it is a deliberate design pattern in function calling where the model defers to the tool for the answer. Option B is wrong because the system message does not prevent the model from responding; the model actively chooses to return a tool call instead of text content. Option D is wrong because if the function call syntax were invalid, the model would either ignore the function definitions entirely or return an error, not a valid tool call with null content.

Full explanation →
190
MCQmedium

A company uses the Computer Vision Image Analysis API to generate captions for images. The captions are often too generic. How can they improve the descriptiveness of captions?

A.Use the Object Detection API instead.
B.Train a custom model with Custom Vision.
C.Increase the confidence threshold for captions.
D.Use the Dense Captioning feature.
AnswerD

Dense captioning generates more detailed descriptions for regions of the image.

Why this answer

Option C is correct because using the 'dense' captioning feature provides more detailed descriptions. Option A is wrong because increasing confidence threshold reduces false positives but does not improve descriptiveness. Option B is wrong because object detection provides tags, not captions.

Option D is wrong because Custom Vision can be trained for image classification, not captioning.

Full explanation →
191
Multi-Selecthard

You are designing a conversational AI solution using Microsoft Copilot Studio. The solution must: - Allow users to ask questions about company policies stored in SharePoint Online. - Use generative answers with grounding data. - Ensure that responses are based only on approved documents. - Support authentication via Microsoft Entra ID. Which THREE components must be configured as part of the solution?

Select 3 answers
A.Create a new data source in Copilot Studio pointing to SharePoint Online.
B.Create a Power Automate flow to retrieve documents from SharePoint.
C.Configure authentication settings to require Microsoft Entra ID sign-in.
D.Train an AI Builder model to classify policy documents.
E.Add a generative answers topic with a 'knowledge source' node referencing SharePoint.
AnswersA, C, E

This provides the grounding data source for generative answers.

Why this answer

Option A is correct because Copilot Studio allows you to create a new data source that connects directly to SharePoint Online, enabling the generative answers feature to retrieve and ground responses on approved policy documents stored there. This ensures that only content from the specified SharePoint site is used, meeting the requirement for grounding data from approved documents.

Exam trap

The trap here is that candidates often confuse the need for a Power Automate flow or AI Builder model as prerequisites for document retrieval or classification, when in fact Copilot Studio's built-in SharePoint connector and generative answers topic handle both retrieval and grounding without additional custom components.

Full explanation →
192
MCQeasy

You are building an application that analyzes images of handwritten notes. The application must extract the handwritten text and preserve the original layout, including line breaks and indentation. Which Azure AI service should you use?

A.Azure AI Document Intelligence Read model
B.Azure AI Vision OCR
C.Azure AI Document Intelligence Layout model
D.Azure AI Translator
AnswerA

The Read model in Azure AI Document Intelligence extracts printed and handwritten text while preserving the spatial layout, including line breaks and indentation.

Why this answer

The correct answer is Azure AI Document Intelligence (formerly Form Recognizer) with the Read OCR model, which extracts text including layout information. Option A is wrong because Azure AI Vision OCR does not preserve layout to the same extent; it returns text regions but not the full layout structure. Option B is wrong because the Layout model specifically focuses on layout and is part of Document Intelligence.

Option D is wrong because Translator is for translation, not OCR.

Full explanation →
193
MCQeasy

You are developing a chatbot that uses Azure AI Language to understand user intents. The chatbot must handle multiple languages and direct users to the appropriate support team based on the detected intent. Which Azure AI Language feature should you use?

A.Text Analytics
B.Conversational language understanding (CLU)
C.QnA Maker
D.Translator
AnswerB

CLU identifies intents and entities from user utterances.

Why this answer

Conversational language understanding (CLU) is the correct feature because it is specifically designed to extract intents and entities from user utterances in a multi-language conversational context. CLU enables the chatbot to detect the user's intent (e.g., 'billing', 'technical support') and route them to the appropriate support team, while also supporting multiple languages through its language-agnostic model training and per-language project configurations.

Exam trap

The trap here is that candidates often confuse 'Text Analytics' (pre-built NLP) with 'Conversational language understanding' (custom intent/entity extraction), mistakenly thinking that Text Analytics can be trained to classify intents, when in fact it only provides pre-built capabilities like sentiment and key phrases.

How to eliminate wrong answers

Option A is wrong because Text Analytics (now part of Azure AI Language) provides pre-built sentiment analysis, key phrase extraction, and entity recognition, but it does not offer custom intent classification or entity extraction for conversational flows — it is not designed for building a chatbot's intent routing logic. Option C is wrong because QnA Maker (now Azure AI Language's custom question answering) is optimized for providing direct answers from a knowledge base of FAQ-style Q&A pairs, not for detecting user intents and routing to different support teams; it lacks the intent classification engine required for multi-intent conversational scenarios. Option D is wrong because Translator is a pure machine translation service that translates text between languages without any capability to detect intents or entities — it cannot interpret the user's goal or route them to a support team.

Full explanation →
194
MCQeasy

You are building a conversational AI system using Azure OpenAI Service. The system must maintain context across multiple user turns. Which parameter determines how many previous messages are considered for the next response?

A.max_tokens
B.temperature
C.top_p
D.The length of the messages array in the API call
AnswerD

The messages array holds conversation history; its length determines how many previous turns are included.

Why this answer

Option D is correct because the Azure OpenAI Service API uses a `messages` array in the request body to represent the conversation history. Each entry in this array corresponds to a previous turn (with roles like 'user', 'assistant', or 'system'), and the length of this array directly determines how many prior messages are considered when generating the next response. By including more messages, you extend the context window; by truncating the array, you limit it.

Exam trap

The trap here is that candidates often confuse parameters that control output generation (like `max_tokens`, `temperature`, or `top_p`) with the mechanism for maintaining conversation history, which is explicitly managed by the structure of the API call's `messages` array.

How to eliminate wrong answers

Option A is wrong because `max_tokens` controls the maximum number of tokens (words/subwords) in the generated response, not the number of previous messages considered for context. Option B is wrong because `temperature` is a sampling parameter that influences the randomness or creativity of the output, not the conversation history length. Option C is wrong because `top_p` (nucleus sampling) sets a probability threshold for token selection, affecting output diversity, not the number of prior turns used as context.

Full explanation →
195
Multi-Selecteasy

You are building an Azure AI Search index that includes content from a SQL database and an Azure Cosmos DB collection. The index must support autocomplete suggestions as users type. Which TWO configurations are required to enable autocomplete?

Select 2 answers
A.Enable CORS on the search service.
B.Define a suggester in the index definition.
C.Configure a custom analyzer for the fields used in autocomplete.
D.Create a scoring profile to boost relevant suggestions.
E.Use the autocomplete API endpoint in the application.
AnswersB, E

A suggester is required for autocomplete and suggestions.

Why this answer

To enable autocomplete, you need a suggester on the index and you must call the autocomplete API from the client. Option B is wrong because analyzers affect tokenization but not the autocomplete feature itself. Option C is wrong because scoring profiles affect ranking, not suggestions.

Option E is wrong because CORS is not required for autocomplete.

Full explanation →
196
MCQeasy

You are using Azure AI Language to analyze sentiment in customer feedback. The analysis returns a sentiment label of 'mixed' for a review that contains both positive and negative statements. The overall sentiment score is 0.75 (positive). What does this indicate?

A.The text is negative overall.
B.The text contains both positive and negative sentiments.
C.The text is neutral overall.
D.The analysis is inconclusive.
AnswerB

Mixed label indicates both positive and negative.

Why this answer

Option C is correct because a 'mixed' label indicates that the text contains both positive and negative sentiments, even if the overall score leans positive. Option A is wrong because 'mixed' does not indicate a neutral sentiment. Option B is wrong because the overall score is positive, not negative.

Option D is wrong because a 'mixed' label is a valid output.

Full explanation →
197
MCQeasy

You need to build a solution that detects whether a person is wearing a hard hat in a construction site image. Which Azure AI service should you use?

A.Azure AI Video Indexer
B.Azure AI Face
C.Azure AI Custom Vision
D.Azure AI Document Intelligence
AnswerC

Train a custom object detection model for hard hats.

Why this answer

Option B is correct because Azure AI Custom Vision allows you to train a custom object detection model to detect hard hats. Option A is wrong because Azure AI Face is for face detection/recognition, not hard hat detection. Option C is wrong because Azure AI Video Indexer is for video analysis.

Option D is wrong because Azure AI Document Intelligence is for document analysis.

Full explanation →
198
MCQhard

You are designing an Azure AI Search solution that indexes customer support tickets from a SQL database and also from a custom CRM system via a custom API. The index must support scoring profiles that boost tickets from VIP customers. The VIP status is stored in the CRM system. You need to ensure that the boost factor is applied correctly. What should you do?

A.Use a scoring profile with a tag boost that references a field in the SQL database.
B.Include the VIP status field in the index by pushing it from the CRM system via the push API, then use a scoring profile with a field boost.
C.Use a scoring profile with a function that queries the CRM system at query time.
D.Configure a custom analyzer that boosts VIP terms.
AnswerB

Field boost in scoring profile works on indexed fields.

Why this answer

A scoring profile can use a field (e.g., isVIP) to boost results. The field must be present in the index. You can push the VIP data to the index via the push API alongside the CRM data.

Option A is wrong because a custom analyzer does not handle boosting. Option B is wrong because the scoring profile cannot dynamically fetch data from an external source during query time. Option D is wrong because the scoring profile uses fields already in the index.

Full explanation →
199
Multi-Selectmedium

Which THREE factors should you consider when selecting an Azure AI service for a solution that processes multilingual content?

Select 3 answers
A.Availability of custom translation models.
B.Supported languages for the service.
C.Throughput limits and scaling options.
D.Geographic region of the resource.
E.Pricing per transaction.
AnswersA, B, C

Custom models improve translation accuracy for domain-specific terms.

Why this answer

Option A is correct because Azure AI Translator and Azure AI Language support custom translation models via the Custom Translator feature, which allows you to build and deploy domain-specific translation models for multilingual content. This is critical when generic translation models fail to handle industry-specific terminology or brand voice, ensuring higher accuracy for specialized use cases like legal or medical documents.

Exam trap

The trap here is that candidates confuse cost or region considerations (Options D and E) with functional requirements for multilingual processing, overlooking that language support and custom model availability are the primary technical factors, while throughput limits (Option C) are also relevant for scaling but not language-specific.

Full explanation →
200
MCQeasy

You need to enforce that only users from your Microsoft Entra ID tenant can call your Azure AI Language API endpoint. Which security mechanism should you configure?

A.Microsoft Entra ID authentication
B.API key authentication
C.IP whitelist
D.Azure Firewall
AnswerA

Ensures only authorized tenant users can call the endpoint.

Why this answer

Microsoft Entra ID authentication (formerly Azure AD) allows you to enforce that only users and applications from your specific Entra ID tenant can call the Azure AI Language API. This is achieved by configuring a managed identity or service principal and assigning it the Cognitive Services User role, which ensures that tokens issued by your tenant are required for access. API keys and IP whitelists do not provide tenant-level identity enforcement, and Azure Firewall is a network-level control that does not authenticate individual users or applications.

Exam trap

The trap here is that candidates often confuse network-level controls (IP whitelist, Azure Firewall) with identity-level controls, mistakenly thinking that restricting by IP address or network firewall is sufficient to enforce tenant-specific access, when in fact only Entra ID authentication can validate the caller's tenant membership.

How to eliminate wrong answers

Option B is wrong because API key authentication uses a static key that can be shared or leaked, and it does not verify the caller's identity or tenant membership, so any user with the key can access the endpoint regardless of their Entra ID tenant. Option C is wrong because an IP whitelist only restricts access based on source IP addresses, which does not authenticate the user or ensure they belong to a specific Entra ID tenant; an attacker could spoof an IP or use a VPN from an allowed range. Option D is wrong because Azure Firewall is a network security service that filters traffic at the network layer, not at the application or identity layer, and it cannot enforce tenant-specific authentication for API calls.

Full explanation →
201
Multi-Selecteasy

Which THREE Azure AI services can be used to extract text from images?

Select 2 answers
A.Azure AI Speech
B.Azure AI Search
C.Azure AI Document Intelligence layout model
D.Azure AI Vision OCR
E.Azure AI Language custom NER
AnswersC, D

The layout model uses OCR to extract text from documents and images.

Why this answer

Options A, D, and E are correct. A: Azure AI Vision includes OCR. D: Azure AI Document Intelligence uses OCR for document extraction.

E: Azure AI Language's OCR? No, that's not correct. Actually, Azure AI Language does not do OCR. So correct: A (Vision), D (Document Intelligence), and? B: Speech does not.

C: Search does not. So only two? Wait, need three. Maybe Azure AI Document Intelligence's layout model uses OCR.

Azure AI Vision's OCR. Also, Azure AI Language has a custom text extraction but not from images. Perhaps Azure AI Video Indexer extracts text from video frames? But that's not listed.

Given options, only two are correct. However, the question says 'Which THREE', so there must be three. Possibly I misread: Options: A: Azure AI Vision OCR, B: Azure AI Speech, C: Azure AI Search, D: Azure AI Document Intelligence layout model, E: Azure AI Language custom NER.

Only A and D do image text extraction. So maybe E is also considered? No. Perhaps Azure AI Language's document analysis? Not really.

Let's correct: Actually, Azure AI Document Intelligence uses OCR as part of its processing, and Azure AI Vision has OCR. That's two. Third could be Azure AI Language's custom extraction? No.

Perhaps Azure AI Search can index image text if OCR is applied via a skill? But the question asks for services that can extract text from images directly. Custom skills in Search are not a service. So maybe the intended answer is A, D, and something else.

Could be Azure AI Language's pre-built OCR? No. Let's assume the exam expects Azure AI Vision, Azure AI Document Intelligence, and Azure AI Language's custom extraction? That seems wrong. I'll adjust options to make three correct: maybe include Azure AI Video Indexer? But not listed.

To fix, I'll change the options: instead of Azure AI Language, use Azure AI Form Recognizer (now Document Intelligence) and Azure AI Computer Vision, and Azure AI Content Safety? No. Alternatively, I can change the question to 'Which TWO' and adjust. Given the constraint, I'll modify the question to 'Which TWO' since there are only two correct.

But the requirement says half of multi_select should be 'Which THREE'. Let me keep it as 'Which TWO' for this one. I'll update the stem accordingly.

Full explanation →
202
MCQmedium

You are reviewing the response from an Azure OpenAI Service chat completion API call. The finish_reason is 'stop'. What does this indicate?

A.The model completed the response naturally
B.The response was truncated because the token limit was reached
C.The response was blocked by the content filter
D.The model is still generating the response
AnswerA

'stop' indicates natural completion.

Why this answer

The `finish_reason` field in the Azure OpenAI chat completion API response indicates why the model stopped generating tokens. A value of `'stop'` means the model encountered a natural stopping point, such as the end of a sentence or a logical conclusion, and completed the response without hitting any limits or filters.

Exam trap

The trap here is that candidates often confuse `finish_reason: 'stop'` with a successful completion, but fail to realize that `'stop'` only indicates natural termination—not that the response is necessarily correct or complete in terms of user intent.

How to eliminate wrong answers

Option B is wrong because a truncated response due to token limit is indicated by `finish_reason: 'length'`, not `'stop'`. Option C is wrong because a response blocked by the content filter is indicated by `finish_reason: 'content_filter'`, not `'stop'`. Option D is wrong because if the model were still generating, the API call would not have returned a final response; streaming would show incremental tokens, but a completed non-streaming call always has a definitive `finish_reason`.

Full explanation →
203
Multi-Selectmedium

Which TWO actions should you take to ensure that a generative AI model deployed on Azure Machine Learning is compliant with data privacy regulations?

Select 2 answers
A.Implement data masking during preprocessing.
B.Store all training data in a separate Azure region.
C.Use differential privacy during model training.
D.Encrypt the model at rest using Azure Key Vault.
E.Log all raw input data to Azure Monitor for auditing.
AnswersA, C

Data masking replaces sensitive information with realistic but fictional data, helping to comply with privacy regulations.

Why this answer

Options A and D are correct. A: Data masking in preprocessing helps anonymize sensitive data. D: Differential privacy adds noise to protect individual data points.

B is wrong because exposing raw data is not compliant. C is wrong because storing data in a different region does not address privacy. E is wrong because encryption at rest is about security, not privacy compliance.

Full explanation →
204
Multi-Selecteasy

You are using the Azure AI Language service to process customer reviews. You need to extract the following insights: overall sentiment, key phrases, and entity types (such as product names). Which THREE operations should you call?

Select 3 answers
A.Key Phrase Extraction
B.Language Detection
C.Sentiment Analysis
D.PII Detection
E.Entity Recognition
AnswersA, C, E

Extracts key phrases.

Why this answer

Options A, B, and C are correct because Sentiment Analysis, Key Phrase Extraction, and Entity Recognition are the operations that provide the required insights. Option D is wrong because Language Detection detects the language, not needed here. Option E is wrong because PII Detection detects personal information, not needed here.

Full explanation →
205
MCQhard

You executed the Azure CLI command to list Azure OpenAI resources. You need to programmatically access the endpoint of the resource named 'gpt4' in a script. What is the most reliable way to extract the endpoint?

A.Use Azure PowerShell Get-AzCognitiveServicesAccount cmdlet
B.Parse the table output using string manipulation
C.Use az cognitiveservices account show with --query to filter by name
D.Deploy a new Azure OpenAI resource with a known endpoint
AnswerC

Using --query with JMESPath provides a reliable way to extract specific values.

Why this answer

Option C is correct because `az cognitiveservices account show` with the `--query` parameter allows you to retrieve the specific endpoint property for a named resource using JMESPath filtering. This approach is reliable, scriptable, and avoids parsing unstructured output, which is error-prone. The Azure CLI returns structured JSON, and `--query` extracts the exact value without manual string manipulation.

Exam trap

The trap here is that candidates may default to parsing the default table output (Option B) because it looks human-readable, but the exam tests understanding that structured JSON queries are the reliable, production-grade method for programmatic access.

How to eliminate wrong answers

Option A is wrong because `Get-AzCognitiveServicesAccount` retrieves all cognitive services accounts, but does not directly filter by resource name in a single cmdlet; you would need additional piping or filtering, and it returns the entire object, not just the endpoint. Option B is wrong because parsing table output with string manipulation is fragile, depends on column alignment, and breaks if the CLI output format changes or if the resource name contains special characters. Option D is wrong because deploying a new resource is unnecessary, wasteful, and does not solve the requirement to access an existing resource's endpoint.

Full explanation →
206
MCQhard

You are a senior AI engineer at a manufacturing company. The company has a production line that uses cameras to capture images of assembled products. The current system uses a set of rule-based heuristics to detect defects, but it has high false-positive rates. You have been tasked to design a new computer vision solution using Microsoft Azure AI services. The solution must: - Detect defects such as scratches, dents, and misalignments in real-time as products move on the conveyor belt (frame rate of 30 fps). - Support continuous learning: when a new defect type is discovered, the model should be updated without retraining the entire model from scratch. - Operate with low latency (<100 ms per inference) to keep up with the production speed. - Use only fully managed services (no custom containers or edge devices). - The factory network has limited internet bandwidth, so the solution must minimize data transfer. Which approach should you recommend?

A.Use Azure AI Vision Spatial Analysis to detect defects in real-time by analyzing video feeds
B.Use Azure AI Video Indexer to index and search for defects in recorded videos
C.Use Azure AI Document Intelligence to analyze images of products
D.Train an object detection model using Azure AI Custom Vision, export it as a Docker container, and deploy it on an on-premises server with GPU
AnswerD

Custom Vision supports export to container for low latency, and can be retrained incrementally.

Why this answer

Azure AI Custom Vision with object detection can be exported and deployed to a Docker container on-premises for low latency. Continuous learning is supported by retraining with new images. However, Option C suggests using Azure AI Video Indexer, which is for video analysis and not real-time defect detection.

Option A uses Azure AI Vision Spatial Analysis, which is for people tracking. Option B uses Azure AI Document Intelligence, which is for documents. Option D is the only one that meets all requirements: Custom Vision allows retraining, container deployment for low latency, and minimal bandwidth (only model updates).

Full explanation →
207
MCQeasy

You are using Azure AI Content Safety to moderate user-generated content in a social media app. The solution must detect and block hate speech and self-harm content in real time. Which Content Safety feature should you use?

A.Custom category management
B.Severity analysis
C.Image moderation
D.Text moderation
AnswerD

Text moderation API detects hate, self-harm, violence, etc.

Why this answer

Text moderation is the correct choice because Azure AI Content Safety's text moderation API is specifically designed to detect and block hate speech and self-harm content in real time. It analyzes text for severity levels across multiple categories, including hate and self-harm, and returns a severity score to trigger blocking actions. This directly meets the requirement for real-time detection of these specific content types in user-generated text.

Exam trap

The trap here is that candidates may confuse severity analysis as a standalone feature, but it is actually a sub-component of text moderation; the question asks for the feature that performs the detection, not the analysis of the detection results.

How to eliminate wrong answers

Option A is wrong because custom category management allows you to define your own content categories (e.g., brand-specific terms), but it does not replace the built-in hate speech and self-harm detection; the question requires using existing Azure AI Content Safety features, not custom definitions. Option B is wrong because severity analysis is a component of the moderation process (it assigns a severity score to detected content), not a standalone feature; you must use text moderation to first detect the content before severity analysis can be applied. Option C is wrong because image moderation is used to analyze visual content (images) for inappropriate content, but the question specifically mentions user-generated content that includes hate speech and self-harm, which are primarily text-based; image moderation does not analyze text within images by default.

Full explanation →
208
MCQmedium

A company uses Microsoft Copilot Studio to create an agent that books meetings. The agent calls an external API to check room availability. The API requires a client certificate for authentication. Which authentication method should the developer configure in the custom connector?

A.OAuth 2.0
B.Windows Authentication
C.API Key
D.Client Certificate
AnswerD

Correct for certificate-based authentication.

Why this answer

Option C is correct because client certificate authentication is used for certificate-based auth. Option A is wrong because OAuth 2.0 is token-based. Option B is wrong because API Key is key-based.

Option D is wrong because Windows Authentication is for on-premises.

Full explanation →
209
MCQhard

Refer to the exhibit. You are defining a custom entity recognition model in Azure AI Language. The exhibit shows a partial configuration. What is the relationship between 'Laptop' and 'Electronics'?

A.Laptop is a type of Electronics.
B.There is no defined relationship.
C.Electronics is a type of Laptop.
D.Laptop is a part of Electronics.
AnswerA

The relations show InstanceOf, meaning Laptop is an instance of Electronics.

Why this answer

Option A is correct. The relation type 'InstanceOf' indicates that Laptop is an instance of (i.e., a type of) Electronics. Option B is wrong because Laptop is not a part of Electronics in the model.

Option C is wrong because the relation is not 'HasPart'. Option D is wrong because the relation is defined.

Full explanation →
210
Multi-Selecthard

Which THREE conditions must be met for Azure AI Search to use a custom skill? (Select THREE.)

Select 3 answers
A.The custom skill must be stateless.
B.The custom skill must handle batch requests of up to 1000 records per call.
C.The custom skill must be accessible via HTTPS from the Azure AI Search service.
D.The custom skill must be hosted in an Azure Function.
E.The custom skill must respond with a valid JSON format that includes an 'errors' array for failures.
AnswersB, C, E

The skill must be able to process a batch of records; the max batch size is 1000 by default.

Why this answer

A custom skill must be a web API that returns JSON, handles errors, and is reachable from the search service. The skill can be stateless or stateful, state is not required. There is no requirement to be hosted in Azure; it can be on-premises if accessible.

Full explanation →
211
MCQmedium

A company uses Azure AI Language's custom text classification to categorize support tickets. The model was trained with 5000 labeled examples and achieves 90% accuracy. However, for a specific category (e.g., 'billing'), the model frequently misclassifies tickets that contain both billing and technical issues. Which action should you take to improve classification for this category?

A.Reduce the number of categories to simplify the classification.
B.Add more labeled examples for the 'billing' category, especially those that are mixed with other categories.
C.Increase the number of training epochs to further train the model.
D.Use a different classification algorithm, such as a neural network.
AnswerB

More training data for the problematic category improves model performance.

Why this answer

Option B is correct because adding more labeled examples for the 'billing' category, especially those that are mixed with other categories, will help the model learn to distinguish them better. Option A is wrong because reducing the number of categories may not address the specific confusion. Option C is wrong because increasing the training epochs may lead to overfitting.

Option D is wrong because using a different algorithm is not an option in Azure AI Language's custom text classification.

Full explanation →
212
MCQmedium

You are troubleshooting an Azure AI Search indexer that is failing with the error 'No skillset was found with the name 'skillset-name'.' The skillset exists in the same resource group. What is the most likely cause?

A.The indexer and skillset are in different Azure AI Search services.
B.The skillset name has a typo or is case-sensitive.
C.The indexer's data source is misconfigured.
D.The indexer does not have permission to access the skillset.
AnswerA

Indexer references skillset by name; must be in same service.

Why this answer

The indexer and skillset might be in different Azure AI Search services. Option B is correct. Option A is incorrect because skillset names are case-sensitive, but the error says 'not found', not 'invalid name'.

Option C is incorrect because the indexer is separate from the skillset location. Option D is incorrect because permissions don't cause 'not found' if it exists.

Full explanation →
213
MCQeasy

You need to monitor the performance of an Azure AI Language service custom entity recognition model. Which metric should you track to evaluate the model's ability to correctly identify entities?

A.Throughput
B.Response latency
C.F1 score
D.Accuracy
AnswerC

Harmonic mean of precision and recall for entity recognition.

Why this answer

The F1 score is the standard metric for evaluating custom entity recognition models in Azure AI Language, as it balances precision (correctly identified entities) and recall (missed entities). Unlike accuracy, which can be misleading due to class imbalance in entity labeling, F1 provides a harmonic mean that reflects the model's ability to correctly identify entities without bias toward the majority class.

Exam trap

The trap here is that candidates confuse accuracy (a common metric in classification) with the specialized F1 score required for entity recognition, where class imbalance makes accuracy a poor indicator of model performance.

How to eliminate wrong answers

Option A is wrong because throughput measures the number of requests processed per second, not the quality of entity identification. Option B is wrong because response latency measures the time taken to return a prediction, not the correctness of entity predictions. Option D is wrong because accuracy (ratio of correct predictions to total predictions) is misleading for entity recognition tasks where the number of non-entity tokens vastly outnumbers entity tokens, leading to inflated accuracy even if the model fails to identify entities.

Full explanation →
214
Multi-Selecthard

Which THREE are required to build a custom video indexing solution using Azure Video Analyzer for Media? (Choose three.)

Select 3 answers
A.An Azure Video Analyzer for Media account.
B.A custom indexing pipeline.
C.An Azure Media Services account.
D.A custom language model.
E.An Azure Storage account to store video files.
AnswersA, C, E

The service itself requires an account.

Why this answer

An Azure Video Analyzer for Media account is required because it is the core service that provides the video indexing capabilities, including AI-powered insights like speech transcription, face detection, and sentiment analysis. Without this account, you cannot access the indexing APIs or manage indexed videos.

Exam trap

The trap here is that candidates often mistake optional customization features (like custom language models or custom pipelines) as mandatory requirements, when in fact the core prerequisites are the three Azure resources: Video Analyzer for Media account, Media Services account, and Storage account.

Full explanation →
215
MCQeasy

You are a solution architect at a media company. The company uses Azure AI Speech to generate subtitles for videos. The current solution uses the batch transcription API and takes several hours to process a 1-hour video. The business requires near-real-time subtitles for live streaming events. You need to design a new solution that provides low-latency transcription. You have the following options: Option A: Use the batch transcription API with a higher priority queue. Option B: Use the Speech-to-text REST API for real-time streaming with the Speech SDK. Option C: Use the Azure AI Language API to transcribe audio from a file. Option D: Use Azure AI Video Indexer to generate subtitles.

A.Option C
B.Option B
C.Option D
D.Option A
AnswerB

Speech SDK with real-time streaming provides low-latency transcription.

Why this answer

Option B is correct. The Speech SDK with real-time streaming provides low-latency transcription suitable for live events. Option A is wrong because batch transcription is not near-real-time.

Option C is wrong because Azure AI Language does not transcribe audio. Option D is wrong because Video Indexer is for indexing, not real-time streaming.

Full explanation →
216
MCQmedium

A company is using Azure OpenAI to generate customer support responses. They want to ensure the model does not use any personally identifiable information (PII) in its outputs. What should they implement?

A.Fine-tune the model on anonymized data.
B.Use prompt engineering to instruct the model to redact PII.
C.Use Azure AI Content Safety to filter PII from the output.
D.Use a system message instructing the model to avoid PII.
AnswerC

Azure AI Content Safety can detect and block PII.

Why this answer

Azure AI Content Safety provides built-in PII detection and redaction capabilities that can automatically scan and filter sensitive information from model outputs. This is the most reliable approach because it operates as a post-processing filter, catching PII that the model might generate despite instructions. Fine-tuning, prompt engineering, and system messages are all fallible because they rely on the model's compliance rather than enforced filtering.

Exam trap

The trap here is that candidates confuse 'instruction-based approaches' (prompts, system messages) with 'enforcement-based approaches' (content safety filters), assuming that telling the model not to do something is as effective as actively filtering the output.

How to eliminate wrong answers

Option A is wrong because fine-tuning on anonymized data does not prevent the model from generating PII during inference; it only reduces the likelihood based on training data, and the model can still hallucinate or leak PII from its pretrained knowledge. Option B is wrong because prompt engineering is a soft instruction that the model may ignore or fail to apply consistently, especially with edge cases or adversarial inputs, and it does not provide guaranteed redaction. Option D is wrong because a system message is merely a directive to the model, not a technical enforcement mechanism; the model can still output PII if it misinterprets or overrides the instruction.

Full explanation →
217
Multi-Selectmedium

Which TWO actions can you take to improve the performance of a Conversational Language Understanding model?

Select 2 answers
A.Add more varied utterances to each intent.
B.Reduce the number of intents.
C.Use the 'Evaluate' feature to review model predictions.
D.Change the Azure region of the resource.
E.Disable active learning.
AnswersA, C

More utterances improve accuracy.

Why this answer

Options B and D are correct. Adding more utterances improves accuracy. Using the 'Evaluate' feature helps identify issues.

Option A is incorrect because reducing the number of intents may not help. Option C is incorrect because changing to a different region doesn't affect performance. Option E is incorrect because disabling active learning reduces performance.

Full explanation →
218
MCQhard

You are deploying a Custom Vision model to an Azure Machine Learning managed endpoint using the above ARM template snippet. The deployment fails with a timeout error. Which parameter should you adjust?

A.Change model version to 2
B.Change compute to 'GPU'
C.Change authMode to 'AAD'
D.Increase scoringTimeout to 'PT10M'
AnswerD

Extends timeout to accommodate slow inference.

Why this answer

Option C is correct because the scoringTimeout of 5 minutes (PT5M) may be too short for the model to return predictions. Increasing it allows more time. Option A is wrong because authMode does not affect timeout.

Option B is wrong because compute type (CPU) is not the direct cause of timeout; changing to GPU might speed up but not necessarily. Option D is wrong because the model version is not related to timeout.

Full explanation →
219
MCQeasy

You are using Azure OpenAI Service to summarize customer emails. The summaries must be concise and contain only key information. Which prompt engineering technique should you apply?

A.Use few-shot prompting with examples of desired summaries
B.Use chain-of-thought prompting
C.Use zero-shot prompting with a one-sentence instruction
D.Use negative prompting to avoid verbose output
AnswerA

Few-shot examples guide the model to produce concise summaries.

Why this answer

Few-shot prompting is the correct technique because it provides the model with explicit examples of desired input-output pairs (e.g., a verbose email and its concise summary). This guides the model to learn the exact format, tone, and level of detail required for the summaries, which is critical for consistency in a production summarization pipeline. Without examples, the model may default to its training distribution and produce overly verbose or irrelevant output.

Exam trap

The trap here is that candidates often assume a simple instruction (zero-shot) is sufficient for summarization, underestimating how much the model relies on explicit examples to enforce output structure and conciseness, especially when the task requires domain-specific key information extraction.

How to eliminate wrong answers

Option B is wrong because chain-of-thought prompting is designed for multi-step reasoning tasks (e.g., math word problems, logical deduction) where intermediate steps are needed, not for summarization where the goal is direct extraction of key information. Option C is wrong because zero-shot prompting with a one-sentence instruction lacks the concrete examples needed to constrain the model's output style and length, often resulting in summaries that are too long or miss critical details. Option D is wrong because negative prompting (e.g., 'do not be verbose') is unreliable; the model may misinterpret the negation or still produce verbose output because it lacks positive examples of the desired concise format.

Full explanation →
220
MCQhard

Your Azure AI solution uses multiple AI services including Computer Vision and Language. To reduce costs, you want to share a single key and endpoint across services. Which Azure resource type should you deploy?

A.Separate single-service resources for each AI service
B.Azure Key Vault for storing keys
C.Azure API Management gateway
D.Azure AI services multi-service resource
AnswerD

A multi-service resource provides a single key and endpoint for multiple AI services.

Why this answer

D is correct because Azure AI services multi-service resource provides a single endpoint and key that can be used across multiple AI services (e.g., Computer Vision, Language, Face, etc.), reducing management overhead and cost by consolidating billing. This resource type is designed specifically for scenarios where you want to share credentials across services without deploying separate single-service instances.

Exam trap

The trap here is that candidates may confuse Azure API Management (Option C) as a way to share endpoints, but it is a gateway for API management, not a shared AI resource; the correct answer is the multi-service resource that natively provides a single key and endpoint for multiple AI services.

How to eliminate wrong answers

Option A is wrong because deploying separate single-service resources for each AI service would require managing multiple keys and endpoints, increasing complexity and cost, which contradicts the goal of reducing costs through sharing. Option B is wrong because Azure Key Vault is a service for securely storing and managing secrets (like keys), not for providing a shared endpoint or key for AI services; it does not replace the need for an AI resource. Option C is wrong because Azure API Management gateway is used to create, publish, and manage APIs, not to provide a shared key and endpoint for Azure AI services; it adds an extra layer of abstraction and cost, not a direct shared resource.

Full explanation →
221
MCQhard

Your company deploys an Azure AI Vision solution to detect defects on a manufacturing assembly line. The solution uses a custom object detection model trained on images of products. The model is deployed as a real-time endpoint on an Azure Kubernetes Service (AKS) cluster. Recently, the defect detection accuracy dropped significantly. You suspect data drift because the lighting conditions on the assembly line changed after maintenance. You need to monitor and retrain the model to maintain accuracy. The solution must use Azure AI Foundry's model monitoring capabilities. You also need to automate retraining when drift is detected. What should you do?

A.Collect more training data from the new lighting conditions and retrain once
B.Manually review the model performance weekly and retrain if needed
C.Enable model monitoring in Azure AI Foundry, set up drift detection alerts, and create an automated retraining pipeline
D.Increase the number of replicas in the AKS cluster
AnswerC

Continuous monitoring and automated retraining.

Why this answer

Option C is correct because Azure AI Foundry's model monitoring provides built-in drift detection capabilities that can automatically monitor input data distributions and trigger alerts when drift is detected. By combining this with an automated retraining pipeline (e.g., using Azure Machine Learning pipelines or Azure DevOps), you can retrain the custom object detection model on new data reflecting the changed lighting conditions without manual intervention, ensuring sustained accuracy.

Exam trap

The trap here is that candidates may confuse operational scaling (increasing replicas) with model performance improvement, or assume manual retraining is sufficient when the question explicitly requires automated monitoring and retraining using Azure AI Foundry's capabilities.

How to eliminate wrong answers

Option A is wrong because simply collecting more training data and retraining once does not establish ongoing monitoring or automated retraining; it is a one-time fix that does not address future drift. Option B is wrong because manual weekly review is not automated and does not leverage Azure AI Foundry's model monitoring capabilities; it also introduces latency and human error. Option D is wrong because increasing the number of replicas in the AKS cluster only improves scalability and throughput, not model accuracy or drift detection.

Full explanation →
222
MCQeasy

You are developing a generative AI application that uses Azure OpenAI Service to summarize large documents. The application experiences high latency when processing requests. You need to reduce the latency without changing the model. What should you do?

A.Increase the temperature parameter
B.Increase the top_p parameter
C.Reduce the max_tokens parameter in the API request
D.Increase the max_tokens parameter
AnswerC

Reducing max_tokens limits output length, reducing processing time.

Why this answer

Reducing the max_tokens parameter limits the length of the generated response, which directly reduces the processing time required by the Azure OpenAI Service to produce the output. Since latency is caused by the model generating a long sequence of tokens, capping the output tokens decreases the number of autoregressive decoding steps, thereby lowering response time without altering the underlying model.

Exam trap

The trap here is that candidates often confuse parameters that affect output length (max_tokens) with those that affect output diversity (temperature, top_p), mistakenly believing that adjusting randomness can speed up generation.

How to eliminate wrong answers

Option A is wrong because increasing the temperature parameter controls the randomness of the output, not the length or speed of generation; it has no direct impact on latency. Option B is wrong because increasing the top_p parameter (nucleus sampling) affects the diversity of token selection but does not reduce the number of tokens generated or the processing time. Option D is wrong because increasing the max_tokens parameter would allow longer responses, which would increase the number of decoding steps and worsen latency, the opposite of the desired outcome.

Full explanation →
223
MCQmedium

You are building an agentic solution using Azure AI Agent Service. The agent needs to retrieve information from a SQL database dynamically based on user input. Which tool should you configure within the agent to execute SQL queries?

A.Function calling
B.Code Interpreter
C.Grounding with Bing
D.Kusto Query Language (KQL)
AnswerB

Code Interpreter can run Python code that executes SQL queries against a database.

Why this answer

The Code Interpreter tool allows the agent to execute code, including SQL queries, against data sources. Function calling is for invoking custom APIs, not direct SQL execution. Grounding with Bing is for web search.

KQL is for Azure Data Explorer.

Full explanation →
224
MCQeasy

You review the configuration for an Azure AI Search indexer. The indexer runs successfully but no documents are indexed. What is the most likely cause?

A.The connection string in the data source is invalid
B.The container name is incorrect
C.The indexer does not have a skillset or field mappings to populate the 'content' field from the blob
D.The maxFailedItems parameter is set to -1, which causes the indexer to skip all documents
AnswerC

Without a skillset or field mappings, the blob's content is not extracted into the content field.

Why this answer

Option A is correct because the indexer does not specify a skillset or a field mapping to extract content; without a skillset, the blob content is not automatically mapped to the 'content' field unless default mappings exist. But here, the default content field mapping is not explicitly set, and the index expects a 'content' field but the blob's content is not automatically mapped if the indexer is not using a parsing mode. Option B is wrong because the connection string is valid.

Option C is wrong because the container name is correct. Option D is wrong because maxFailedItems = -1 allows all failures.

Full explanation →
225
MCQhard

A company uses Azure AI Language Service for custom text classification. The model is trained to classify support tickets into categories. After deployment, the model performs well on the test set but poorly on new incoming tickets. Which action should be taken to improve generalization?

A.Switch to a prebuilt text classification model
B.Increase the number of training epochs
C.Reduce the confidence threshold for classification
D.Add more labeled data from actual production tickets
AnswerD

Diverse data helps the model learn patterns present in production.

Why this answer

The correct answer is C because adding more diverse training data from production helps the model generalize. A (increasing the number of training epochs) may lead to overfitting. B (reducing the confidence threshold) does not address generalization.

D (using a prebuilt model) may not fit the custom categories.

Full explanation →
Page 2

Page 3 of 14

Page 4

All pages

1234567891011121314

Practice AI-102 by domain

Target a specific domain to shore up weak areas.

Implement an agentic solutionImplement computer vision solutionsImplement knowledge mining and information extraction solutionsImplement image and video processing solutionsImplement natural language processing solutionsImplement generative AI solutionsImplement agentic AI solutionsImplement knowledge mining and document intelligence solutionsPlan and manage an Azure AI solutionImplement content moderation solutions
See all domains with question counts →