Courseiva

Microsoft Azure AI Engineer Associate AI-102 (AI-102) — Questions 76–150

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

Take a mock examExam hub
Page 1

Page 2 of 14

Page 3
76
MCQhard

A financial services company is building an agent that uses Azure OpenAI to generate investment advice. The agent must be monitored for toxicity and bias. Which combination of services should the team use to implement content safety monitoring?

A.Azure Cognitive Search and Azure AI Language.
B.Azure Bot Service and Azure Logic Apps.
C.Azure Machine Learning and Azure Functions.
D.Azure AI Content Safety and Azure OpenAI content filtering.
AnswerD

These services provide comprehensive content safety.

Why this answer

Azure AI Content Safety provides built-in models for detecting harmful content such as hate speech, self-harm, and sexual content, while Azure OpenAI content filtering applies configurable severity-level filters (e.g., low, medium, high) to model inputs and outputs. Together, they enable real-time monitoring of toxicity and bias in generated investment advice, meeting compliance requirements for financial services.

Exam trap

The trap here is that candidates may confuse general AI services (like Azure AI Language or Azure Machine Learning) with the specific, purpose-built content safety and filtering services required for monitoring toxicity and bias in generative AI outputs.

How to eliminate wrong answers

Option A is wrong because Azure Cognitive Search is a retrieval service for indexing and querying data, not a content safety or bias detection tool, and Azure AI Language provides NLP features like sentiment analysis but lacks dedicated toxicity and bias monitoring for generative AI outputs. Option B is wrong because Azure Bot Service is a framework for building conversational agents, and Azure Logic Apps is an integration workflow service; neither includes built-in content safety or bias detection capabilities. Option C is wrong because Azure Machine Learning is a platform for training and deploying custom ML models, and Azure Functions is a serverless compute service; while you could build custom safety logic, they do not provide the pre-built, configurable content filtering and toxicity detection that Azure AI Content Safety and Azure OpenAI content filtering offer out of the box.

Full explanation →
77
Multi-Selectmedium

You are building a generative AI chatbot using Microsoft Copilot Studio. The chatbot must answer questions from a PDF document and a SQL database. Which THREE data sources can you configure? (Choose three.)

Select 3 answers
A.Custom connector to SQL database
B.SharePoint (store the PDF document)
C.Azure AI Search index
D.Dataverse (store SQL data)
E.Azure Blob Storage
AnswersA, B, D

Custom connectors allow integration with SQL databases.

Why this answer

A is correct because Microsoft Copilot Studio supports custom connectors to access external data sources like SQL databases. This allows the chatbot to query the SQL database directly using the connector's API, enabling real-time data retrieval for generative AI responses.

Exam trap

The trap here is that candidates often assume Azure AI Search or Blob Storage are natively configurable in Copilot Studio, but they require additional middleware or custom connectors, unlike SharePoint and Dataverse which are first-party supported sources.

Full explanation →
78
MCQmedium

You see the exhibit from an Azure OpenAI chat completion request. The assistant is not calling the get_weather function when asked about the weather. What is the most likely reason?

A.The required parameter 'location' is missing from the user message
B.Function calling is not supported in the current Azure OpenAI version
C.The function definition is malformed
D.The function is defined in the system message but not in the 'tools' array
AnswerD

Functions must be passed in the 'tools' parameter of the API request.

Why this answer

Option D is correct because the function definition must be included in the 'tools' array of the request, not in the system message. The system message can describe functions, but the actual definition must be in the 'tools' parameter. Option A is wrong because function calling is supported.

Option B is wrong because there is a required parameter. Option C is wrong because the definition looks correct.

Full explanation →
79
MCQhard

Refer to the exhibit. You are calling the Azure AI Language API for conversational language understanding (CLU). The CLU project 'SupportBot' has an intent 'CancelOrder' with an entity 'OrderNumber' of type 'Number'. The deployment 'production' is active. What is the expected output?

A.The response will contain an error because the deployment is not active.
B.The response will contain the entity 'OrderNumber' with value '#12345' because the hash is part of the entity.
C.The response will contain only the top intent, but no entities because the entity type is not recognized.
D.The response will contain the top intent 'CancelOrder' and the entity 'OrderNumber' with value '12345'.
AnswerD

The model correctly identifies intent and entity.

Why this answer

Option D is correct because the CLU model will predict the 'CancelOrder' intent and extract the entity 'OrderNumber' with value '12345'. Option A is wrong because the intent is indeed recognized. Option B is wrong because the entity is extracted.

Option C is wrong because the entity is correctly extracted as a number.

Full explanation →
80
MCQmedium

You have an Azure AI Search skillset defined as shown in the exhibit. When you run the indexer, the enrichment pipeline produces outputs but no entities are extracted. The source documents are in English and contain clear organization and person names. What is the most likely cause?

A.The skill output is not mapped to the index.
B.The skills are in the wrong order.
C.The documents are not in English.
D.The '/document/content' field is an array, but the skill expects a string.
AnswerD

EntityRecognitionSkill expects a string input.

Why this answer

The default version of EntityRecognitionSkill (V3) requires the 'text' input to be a string, not an array. If '/document/content' is an array (e.g., from OCR output), the skill fails silently. Option A is wrong because the language is correct.

Option B is wrong because the skill outputs are configured. Option D is wrong because the skill is in the correct order (no dependency).

Full explanation →
81
MCQhard

Refer to the exhibit. You are troubleshooting an Azure OpenAI API call that is returning incomplete responses. The response stops mid-sentence. Which parameter should you adjust?

A.Increase max_tokens to 1000.
B.Remove the stop parameter.
C.Increase temperature to 1.0.
D.Increase top_p to 1.0.
AnswerA

Increases token budget for response.

Why this answer

The `max_tokens` parameter controls the maximum number of tokens the model can generate in a single response. When a response stops mid-sentence, it typically means the token limit was reached before the model could complete its output. Increasing `max_tokens` to 1000 provides more room for the model to finish its generation, resolving the truncation issue.

Exam trap

The trap here is that candidates confuse parameters that control output length (`max_tokens`) with those that control output diversity (`temperature`, `top_p`) or early stopping (`stop`), leading them to pick options that change style rather than capacity.

How to eliminate wrong answers

Option B is wrong because removing the `stop` parameter would not fix mid-sentence truncation; the `stop` parameter defines sequences that halt generation early, and removing it could actually make responses longer but does not address a hard token limit. Option C is wrong because increasing `temperature` to 1.0 increases randomness and creativity in the output, but does not affect the maximum length of the response; it could even lead to more verbose or erratic completions. Option D is wrong because increasing `top_p` to 1.0 enables nucleus sampling with all tokens considered, which may alter the diversity of the output but does not extend the token budget; the model will still stop when `max_tokens` is exhausted.

Full explanation →
82
MCQhard

Your knowledge mining solution uses Azure AI Search with cognitive skills. During testing, you notice that some documents are not being enriched because the skillset execution fails. Which diagnostic step should you take first?

A.Enable debug mode on the skillset
B.Review the indexer execution history in the portal
C.Re-run the indexer with a fresh document
D.Check the indexer logs in Azure Monitor
AnswerA

Debug mode provides detailed per-document skill execution logs.

Why this answer

Option B is correct because enabling debug mode creates a debug session that captures detailed execution logs for each skill. Option A is wrong because checking the indexer status shows success/failure but not detailed skill errors. Option C is wrong because resubmitting might repeat the same error without insight.

Option D is wrong because the portal shows indexer execution history but not per-document skill details.

Full explanation →
83
MCQmedium

Your company is deploying an Azure AI Document Intelligence solution to process invoices. The solution must: - Extract key fields (invoice number, date, total amount). - Handle invoices in both PDF and image formats. - Use a prebuilt model to reduce development effort. - Process high volumes (up to 10,000 invoices per day). - Store extracted data in Azure Cosmos DB. You need to design the processing pipeline. What should you do?

A.Use the synchronous API of the prebuilt invoice model. For each invoice, call the API and write the result to Cosmos DB.
B.Use the prebuilt invoice model with the async API. Submit all invoices for analysis. Use an Azure Function to poll for results and write to Cosmos DB.
C.Use the prebuilt receipt model to process invoices. Store results in Cosmos DB.
D.Use the layout model to extract text from invoices. Then use Azure AI Language to extract entities.
AnswerB

Async API handles high volume; Azure Function automates the workflow.

Why this answer

Option B is correct because the prebuilt invoice model's asynchronous API is designed for high-volume batch processing, allowing you to submit up to 10,000 invoices per day without timeout or rate-limit issues. The async API returns operation locations that you can poll via an Azure Function, and once results are ready, you write the extracted fields (invoice number, date, total amount) to Azure Cosmos DB. This decouples submission from retrieval, ensuring scalability and reliability for both PDF and image formats.

Exam trap

The trap here is that candidates assume the synchronous API is simpler and sufficient for high volume, but they overlook the rate limits and payload constraints that make the async API mandatory for production-scale invoice processing.

How to eliminate wrong answers

Option A is wrong because the synchronous API has strict payload size and rate limits (e.g., 8 MB per request, 15 requests per second per region), making it unsuitable for processing 10,000 invoices daily without throttling or timeouts. Option C is wrong because the prebuilt receipt model is trained on receipts, not invoices, so it will fail to extract invoice-specific fields like invoice number and total amount accurately. Option D is wrong because using the layout model plus Azure AI Language for entity extraction is a custom, multi-step approach that increases development effort and complexity, contradicting the requirement to use a prebuilt model to reduce development effort.

Full explanation →
84
MCQeasy

You need to extract key-value pairs from a large set of invoices. The invoices have a consistent layout but vary in format (PDF, TIFF). Which Document Intelligence model should you use?

A.Custom extraction model
B.Layout model
C.Read model
D.Premade invoice model
AnswerD

Built specifically for invoices.

Why this answer

Option B is correct because the premade invoice model is designed for common invoice layouts. Option A is wrong because the layout model extracts text and tables, not key-value pairs. Option C is wrong because custom extraction requires training data.

Option D is wrong because the read model only extracts text.

Full explanation →
85
MCQhard

You are deploying a custom named entity recognition (NER) model using Azure AI Language. The model must extract product codes that follow a specific pattern (e.g., 'PRD-12345'). You have 5,000 labeled examples. After training, the model extractor works well on development data but fails to extract product codes from new data. What is the most likely issue?

A.The training data size is insufficient.
B.The product code pattern is too complex for the model to learn.
C.The model is overfitting to the training data.
D.The labeling is inconsistent across the dataset.
AnswerC

Overfitting causes good performance on training data but poor on new data.

Why this answer

Option D is correct because the model likely overfits to the training data, failing to generalize to variations. Option A is wrong because insufficient training data would cause poor performance on development data too. Option B is wrong because labeling consistency issues would affect training and development performance.

Option C is wrong because the model is designed to extract patterns, but overfitting is the issue.

Full explanation →
86
MCQmedium

A company uses Azure AI Language Service to analyze customer feedback. They notice that the sentiment scores for negative reviews are often incorrectly labeled as neutral. Which configuration should be adjusted to improve accuracy?

A.Deploy the Language service in a different Azure region
B.Increase the confidence threshold for sentiment classification
C.Create a custom sentiment analysis model using Custom Text Classification
D.Enable Key Phrase Extraction to preprocess the text
AnswerC

A custom model can be trained on domain-specific data to improve accuracy.

Why this answer

The correct answer is C because sentiment analysis accuracy can be improved by providing a custom model trained on domain-specific data. A (increasing confidence threshold) would reduce false positives but not address mislabeling. B (changing to a different region) does not affect model accuracy.

D (using key phrase extraction) is unrelated to sentiment.

Full explanation →
87
MCQeasy

You are building a solution to detect if a person is wearing a hard hat in construction site images. You have a small dataset of labeled images. Which Azure service should you use?

A.Azure AI Vision Image Analysis
B.Azure Video Indexer
C.Azure AI Document Intelligence
D.Azure Custom Vision
AnswerD

Custom Vision trains on your labeled images.

Why this answer

Option B is correct because Custom Vision allows you to train a custom object detection model with your own images. Option A is wrong because Azure AI Vision prebuilt models do not include hard hat detection. Option C is wrong because Video Indexer is for video analysis, not static images.

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

Full explanation →
88
MCQeasy

You need to extract entities such as dates, locations, and organization names from unstructured text documents. Which Azure AI service should you use?

A.Computer Vision
B.Azure AI Language Service
C.Azure AI Document Intelligence
D.Azure AI Speech Service
AnswerB

Language Service provides NER capabilities.

Why this answer

Option B is correct because Language Service includes Named Entity Recognition. Option A is wrong because Computer Vision analyzes images. Option C is wrong because Azure AI Document Intelligence extracts structured fields from documents.

Option D is wrong because Speech Service processes audio.

Full explanation →
89
MCQmedium

You are building an application that processes scanned invoices to extract key fields such as total amount, invoice date, and vendor name. The application uses Azure AI Document Intelligence. You need to ensure high accuracy for field extraction without manual labeling. Which feature should you use?

A.Use the General Document model
B.Use the Read API
C.Use a custom neural model
D.Use the Layout API
AnswerC

Custom neural models can be trained on a small set of sample documents to extract specific fields without manual labeling.

Why this answer

Option C is correct because Custom Neural Models are designed to extract fields from structured or semi-structured documents with high accuracy and do not require manual labeling if you use prebuilt models or start with a sample set. Option A is wrong because the Read API only extracts text, not key-value pairs. Option B is wrong because the Layout API extracts text and structure but not custom fields.

Option D is wrong because the General Document model extracts key-value pairs but may not be optimized for invoices without customization.

Full explanation →
90
MCQmedium

You are deploying an Azure AI Document Intelligence solution to process invoices. The solution must extract line-item details such as product code, quantity, and unit price. Which prebuilt model should you use?

A.prebuilt-receipt
B.prebuilt-invoice
C.prebuilt-idDocument
D.prebuilt-layout
AnswerB

Designed for invoice processing with line-item extraction.

Why this answer

The prebuilt-invoice model is specifically designed to extract line-item details such as product code, quantity, and unit price from invoices. It uses deep learning models trained on thousands of invoice samples to identify and extract structured data, including tables and line items, making it the correct choice for this requirement.

Exam trap

The trap here is that candidates might choose prebuilt-layout thinking it can extract any table data, but it lacks the specialized field mapping and labeling that prebuilt-invoice provides for invoice-specific line items.

How to eliminate wrong answers

Option A is wrong because prebuilt-receipt is optimized for receipt documents, focusing on fields like merchant name, transaction date, and total amount, not the detailed line-item structure (product code, quantity, unit price) found in invoices. Option C is wrong because prebuilt-idDocument is designed to extract information from government-issued identification documents (e.g., driver's licenses, passports), such as ID number, name, and date of birth, and has no capability for invoice line-item extraction. Option D is wrong because prebuilt-layout extracts text, tables, and selection marks from documents without specialized field extraction for invoices; it returns raw table data but lacks the pre-trained model logic to identify and label specific invoice fields like product code or unit price.

Full explanation →
91
MCQeasy

Refer to the exhibit. A developer creates an agent in Azure AI Foundry with a code_interpreter tool. The agent is supposed to generate plots but returns errors. What is the most likely cause?

A.The instructions are too generic
B.The JSON syntax is incorrect
C.The code interpreter tool does not support visualization libraries
D.The agent name is invalid
AnswerC

Code interpreter has limited capabilities; plotting may require additional configuration.

Why this answer

The code_interpreter tool in Azure AI Foundry is designed for executing Python code in a sandboxed environment, but it does not include or support visualization libraries such as Matplotlib, Seaborn, or Plotly. Therefore, any attempt to generate plots using these libraries will result in errors, making option C the correct answer.

Exam trap

The trap here is that candidates assume the code_interpreter tool supports all common Python libraries, including visualization ones, because it is based on Python, but Azure AI Foundry intentionally restricts the environment to prevent security risks and resource abuse.

How to eliminate wrong answers

Option A is wrong because generic instructions do not cause runtime errors; they may lead to incorrect outputs but not execution failures. Option B is wrong because JSON syntax errors would be caught during agent creation or configuration, not during plot generation. Option D is wrong because an invalid agent name would prevent agent creation entirely, not cause runtime errors when generating plots.

Full explanation →
92
Multi-Selectmedium

Which TWO Azure AI Search features are used to map skill outputs to search index fields? (Select TWO.)

Select 2 answers
A.indexer parameters
B.outputFieldMappings
C.skillset outputs
D.fieldMappings
E.index fields
AnswersB, E

outputFieldMappings map skill outputs to index fields.

Why this answer

outputFieldMappings map outputs from skills to index fields. Field mappings map source data fields to index fields directly. Skillset outputs are intermediate; indexer outputs are final.

Full explanation →
93
MCQmedium

A retail company uses Azure Computer Vision to analyze in-store camera feeds. They recently added a new product line and updated the object detection model. However, the model fails to detect the new products. What should the company do first?

A.Use the pre-built 'products' model from Computer Vision.
B.Increase the confidence threshold in the API call.
C.Retrain the custom object detection model with images of the new products.
D.Recreate the Computer Vision resource in a different region.
AnswerC

Custom models need retraining with new labeled data.

Why this answer

The model fails to detect new products because it was never trained on them. Retraining the custom object detection model with labeled images of the new products is the correct first step, as it updates the model's knowledge to recognize the new product line. Pre-built models or threshold adjustments cannot add new object classes.

Exam trap

The trap here is that candidates may assume a pre-built model or a simple threshold tweak can handle new object classes, when in fact custom object detection requires retraining with labeled examples of the new items.

How to eliminate wrong answers

Option A is wrong because the pre-built 'products' model from Computer Vision is a fixed, general-purpose model that cannot be extended to recognize custom or newly introduced product lines. Option B is wrong because increasing the confidence threshold would only filter out low-confidence detections, not enable detection of entirely new object classes that the model was never trained to recognize. Option D is wrong because recreating the Computer Vision resource in a different region has no impact on the model's ability to detect new products; region selection affects data residency and latency, not model capabilities.

Full explanation →
94
Multi-Selecteasy

Which TWO capabilities are provided by Azure AI Language's pre-built entity recognition?

Select 2 answers
A.Identifying domain-specific medical terms
B.Identifying names of people
C.Extracting key phrases from text
D.Identifying organization names
E.Determining overall sentiment of the text
AnswersB, D

Pre-built entity recognition includes Person entities.

Why this answer

Options A and D are correct. Pre-built entity recognition can identify people (A) and organizations (D). Option B is wrong because custom entities require custom NER.

Option C is wrong because key phrase extraction is a separate feature. Option E is wrong because sentiment analysis is a separate feature.

Full explanation →
95
MCQeasy

A company is building an agentic solution using Azure AI Agent Service. The agent needs to execute a Power Automate flow when a user requests a vacation approval. Which action type should the developer add to the agent's action definition?

A.httpRequest
B.powerAutomateFlow
C.openApi
D.function
AnswerB

Correct action type for Power Automate flows.

Why this answer

Option D is correct because 'powerAutomateFlow' is the action type for executing Power Automate flows. Option A is wrong because 'openApi' is for REST APIs. Option B is wrong because 'httpRequest' is for direct HTTP calls.

Option C is wrong because 'function' is for custom code functions.

Full explanation →
96
MCQeasy

A team is developing a solution to automatically summarize long documents using Azure AI Language. Which feature should they use?

A.Sentiment analysis.
B.Key phrase extraction.
C.Extractive summarization.
D.Entity recognition.
AnswerC

Extracts key sentences to create a summary.

Why this answer

Option A is correct because the extractive summarization feature in Azure AI Language extracts key sentences from documents. Option B is wrong because key phrase extraction extracts phrases, not a summary. Option C is wrong because sentiment analysis determines sentiment, not summary.

Option D is wrong because entity recognition identifies entities.

Full explanation →
97
MCQhard

Based on the exhibit, which entity should you focus on improving by adding more labeled examples?

A.Date
B.OrderNumber
C.All entities need improvement.
D.ProductName
AnswerD

Low recall (0.65) indicates many ProductName entities are missed.

Why this answer

Option A is correct because ProductName has the lowest recall (0.65) and F1 score (0.76), indicating the model is missing many instances. Adding more positive examples for ProductName would likely improve recall. Option B is wrong because OrderNumber has high recall and F1.

Option C is wrong because Date has high precision and F1. Option D is wrong because all entities may not need improvement; ProductName is the weakest.

Full explanation →
98
MCQhard

You work for a manufacturing company that uses Azure AI services to automate quality inspection on a production line. You have a Custom Vision object detection model that identifies defects on metal parts. The model was trained on images captured under ideal lighting conditions. However, when deployed in the factory, the model's accuracy drops significantly due to inconsistent lighting and glare. You need to improve the model's robustness without collecting new images from the factory floor. What should you do?

A.Increase the number of training iterations to force the model to learn more features.
B.Apply data augmentation techniques such as brightness, contrast, and blur adjustments to the existing training images.
C.Use higher resolution images for training.
D.Change the model type from object detection to classification.
AnswerB

Data augmentation simulates real-world variability and improves generalization.

Why this answer

Option B is correct because using data augmentation techniques like brightness and contrast adjustments, rotation, and noise injection can simulate various lighting conditions and improve robustness. Option A is wrong because increasing training iterations may overfit to the existing data. Option C is wrong because higher resolution does not address lighting variation.

Option D is wrong because changing the model type does not address the data issue.

Full explanation →
99
MCQhard

Your team is implementing a knowledge mining solution using Azure AI Search with custom skills. The custom skill, deployed as an Azure Function, calls a third-party API to enrich documents. You notice that some documents fail enrichment with HTTP 429 (too many requests) errors. You need to ensure all documents are enriched without losing data. What should you do?

A.Configure the custom skill to execute in batch mode and set a retry policy on the indexer
B.Increase the number of partitions in the Azure AI Search service
C.Enable indexer error handling to skip failed documents
D.Scale out the Azure Function to multiple instances
AnswerA

Batch mode reduces API calls, and retry policy handles transient failures.

Why this answer

Option D is correct because configuring the custom skill to execute in batch mode with a retry policy allows the skillset to handle throttling by retrying failed batches. Option A is wrong because increasing the number of partitions does not affect custom skill execution. Option B is wrong because scaling the Azure Function may help but the skillset execution is controlled by the indexer.

Option C is wrong because Azure AI Search manages retries for built-in skills, but custom skills need explicit retry configuration.

Full explanation →
100
MCQmedium

You need to analyze images to detect objects and read text from documents using a single Azure AI service. Which service should you use?

A.Azure AI Vision
B.Azure AI Document Intelligence (Form Recognizer)
C.Azure AI Custom Vision
D.Azure AI Language Service
AnswerA

Azure AI Vision includes both object detection and OCR capabilities.

Why this answer

Azure AI Vision is the correct choice because it provides both image analysis (object detection) and optical character recognition (OCR) for reading text from documents within a single service. Its Read API and Analyze Image API cover both requirements without needing separate services.

Exam trap

The trap here is that candidates often confuse Azure AI Document Intelligence (Form Recognizer) as the only service for text extraction, overlooking that Azure AI Vision also provides OCR for general document text reading.

How to eliminate wrong answers

Option B is wrong because Azure AI Document Intelligence (Form Recognizer) is specialized for extracting structured data from forms and documents, not general object detection in images. Option C is wrong because Azure AI Custom Vision is designed for training custom image classification and object detection models, not for reading text from documents. Option D is wrong because Azure AI Language Service focuses on natural language processing tasks like sentiment analysis and key phrase extraction, not image analysis or OCR.

Full explanation →
101
Multi-Selectmedium

Which TWO actions should you take to ensure that an Azure AI Language Service custom entity recognition model complies with data privacy regulations?

Select 2 answers
A.Use prebuilt entity recognition models instead of custom
B.Increase the number of training epochs
C.Enable diagnostic logging for audit trails
D.Configure data retention policies to delete data after processing
E.Anonymize or remove PII from training data
AnswersD, E

Retention policies ensure data is not stored longer than needed.

Why this answer

The correct answers are B and D. B ensures that training data does not contain PII that could be exposed. D ensures that processed data is not stored longer than necessary.

A (enabling logging) does not directly address privacy. C (using prebuilt models) may not be relevant. E (increasing compute) is unrelated.

Full explanation →
102
MCQeasy

You are designing a solution to extract structured data from a large number of handwritten forms. The forms are scanned and stored as images. Which Azure AI feature should you use?

A.Azure AI Vision's image analysis
B.Azure AI Speech to text
C.Azure Bot Service
D.Azure AI Document Intelligence's OCR capability
AnswerD

OCR extracts text from images, including handwriting.

Why this answer

Option D is correct because the OCR skill in Azure AI Document Intelligence (or Azure AI Search) can extract text from handwritten images. Option A is for speech. Option B is for image analysis, not text extraction from forms.

Option C is for conversational AI.

Full explanation →
103
MCQmedium

You are developing an application that processes images of handwritten forms. The forms contain checkboxes that may be checked or unchecked. Which Azure AI service should you use to detect the state of the checkboxes?

A.Azure AI Custom Vision
B.Azure AI Language
C.Azure AI Document Intelligence
D.Azure AI Computer Vision
AnswerC

Document Intelligence includes trained models for extracting marks from forms.

Why this answer

Option B is correct because Azure AI Document Intelligence (formerly Form Recognizer) is designed to extract information from forms, including checkboxes. The other options are not specialized for checkbox detection: Computer Vision for OCR, Custom Vision for image classification, and Language Service for text analysis.

Full explanation →
104
Matchingmedium

Match each Azure AI concept to its description.

Drag a concept onto its matching description — or click a concept then click the description.

Concepts
Matches

URL to access a cognitive service

Authentication credential for API access

Azure datacenter location for the service

Container for related Azure resources

Billing and access management container

Why these pairings

These are fundamental concepts for managing Azure services.

Full explanation →
105
MCQhard

You have an Azure AI Search solution that indexes customer support tickets. The index includes a 'category' field that should be automatically populated using a custom skill that calls an Azure Machine Learning model. However, the skill fails intermittently with HTTP 429 errors. What is the most likely cause and the best fix?

A.The skill has a timeout set too short; increase the timeout.
B.The skill execution batch size is too large, causing high call volume; reduce batch size in the skillset definition.
C.The indexer runs too frequently; increase the indexer schedule interval.
D.The skill output field mapping is incorrect; fix the mapping.
AnswerB

Reducing batch size decreases calls per execution.

Why this answer

HTTP 429 indicates throttling. The custom skill is being called too frequently. Option B reduces the batch size to lower the call rate.

Option A increases parallel requests, making it worse. Option C changes the skill's input, not the throttling issue. Option D is plausible but less specific than reducing batch size.

Full explanation →
106
MCQhard

A company is using Azure OpenAI Service to power a customer support agent. The agent sometimes generates incorrect information when it cannot find an answer in the knowledge base. The team wants to ensure the agent only responds using information from the knowledge base and explicitly states when it does not know the answer. Which configuration should the team use?

A.Use a custom model fine-tuned on the knowledge base and disable content filtering.
B.Use a system message that says 'If you don't know, say you don't know' and rely on the model's training.
C.Use 'use your own data' feature with strict content filtering and set the model to only respond based on retrieved documents.
D.Use prompt engineering with a system message that instructs the model to only answer from the knowledge base, with no additional filtering.
AnswerC

This ensures responses are grounded in the provided data.

Why this answer

Option C is correct because the 'use your own data' feature in Azure OpenAI Service allows you to restrict the model to answer only from the retrieved documents, ensuring responses are grounded in the knowledge base. Strict content filtering further prevents the model from generating unverified information, and the explicit setting to respond based solely on retrieved documents directly addresses the requirement to state when it does not know the answer.

Exam trap

The trap here is that candidates often confuse prompt engineering (Option D) with a reliable grounding mechanism, not realizing that without a retrieval-augmented generation (RAG) architecture and strict content filtering, the model can still hallucinate even when instructed otherwise.

How to eliminate wrong answers

Option A is wrong because fine-tuning a custom model on the knowledge base does not guarantee the model will only use that knowledge; it can still hallucinate or generate information outside the training data, and disabling content filtering removes safety guardrails without solving the grounding issue. Option B is wrong because relying on a system message and the model's training is insufficient; the model may still generate incorrect information when it cannot find an answer, as it has no mechanism to enforce grounding to the knowledge base. Option D is wrong because prompt engineering alone, even with a system message instructing the model to only answer from the knowledge base, does not provide a technical enforcement mechanism; the model can still hallucinate or generate responses not present in the knowledge base without the retrieval-augmented generation (RAG) architecture that 'use your own data' provides.

Full explanation →
107
MCQmedium

Refer to the exhibit. You submit this request to Azure AI Language's conversational language understanding (CLU) for the 'FlightBooking' project. The model correctly identifies the intent as 'BookFlight' and extracts entities: 'Seattle' as FromCity, 'New York' as ToCity, and 'June 15th' as Date. What is the next step for the application?

A.Call a separate booking API with the extracted entities to complete the reservation.
B.Use the CLU response to directly book the flight via the Azure AI Language service.
C.Prompt the user to rephrase the request because the intent is ambiguous.
D.Send another request to CLU to confirm the booking details.
AnswerA

The application must use the extracted information to call an external API.

Why this answer

Option B is correct. After CLU extracts intent and entities, the application must call a booking API to complete the action. Option A is wrong because the CLU model does not perform actions.

Option C is wrong because the response already contains the entities. Option D is wrong because the intent is already identified.

Full explanation →
108
MCQhard

Refer to the exhibit. You are using Azure AI Search with a skillset that splits documents into pages and then analyzes sentiment per page. You notice that the sentiment analysis is returning unexpected results, such as positive sentiment for negative content. What is the most likely cause?

A.The SentimentSkill is receiving the entire document text instead of individual pages.
B.The split skill context '/document' should be '/document/content'.
C.The defaultLanguageCode is set to 'en' but the documents contain other languages.
D.The page split mode with overlap may still cut sentences, causing sentiment to be evaluated on incomplete sentences.
AnswerD

Splitting at page boundaries can break sentences, leading to inaccurate sentiment analysis.

Why this answer

Option D is correct because the SplitSkill uses 'pages' mode which often splits at page boundaries, potentially cutting sentences in half. This can cause the sentiment analysis to receive incomplete text, leading to incorrect sentiment. Option A is not the primary issue.

Option B is about the initial text, not splitting. Option C is about the split context, but the sentiment context is correct.

Full explanation →
109
MCQmedium

You are implementing a RAG (Retrieval-Augmented Generation) solution using Azure AI Search and Azure OpenAI Service. The solution is returning answers that are not relevant to the user query. What is the most likely cause?

A.The max_tokens parameter is set too high.
B.The chunk size is too small.
C.The index includes too many documents.
D.The relevance score threshold is set too low.
AnswerD

Low threshold includes less relevant documents, leading to poor responses.

Why this answer

Option B is correct because low relevance score thresholds include irrelevant documents in the context. Option A is wrong because chunk size affects granularity but not relevance directly. Option C is wrong because tokens limit affects response length, not relevance.

Option D is wrong because indexing all documents might be desired; the issue is retrieval quality.

Full explanation →
110
MCQeasy

You are planning to use Azure AI Content Safety to moderate user-generated content in a social media application. The solution must detect hate speech and self-harm content. Which Content Safety features should you enable?

A.Severity levels for all categories
B.Hate and self-harm content filters
C.Custom categories for hate speech and self-harm
D.Image moderation
AnswerB

These are built-in categories in Content Safety.

Why this answer

Azure AI Content Safety provides pre-built filters for specific harm categories, including hate speech and self-harm. Enabling the 'Hate and self-harm content filters' directly activates the detection models for these categories, meeting the requirement without needing custom categories or additional features.

Exam trap

The trap here is that candidates might think custom categories are needed for specific harm types like self-harm, but Azure AI Content Safety already includes these as built-in categories, so enabling the pre-built filters is the correct approach.

How to eliminate wrong answers

Option A is wrong because severity levels are a configuration setting within each category filter, not a feature to enable; they adjust sensitivity but don't activate detection for specific categories. Option C is wrong because custom categories are for defining new harm types not covered by built-in filters, but hate speech and self-harm are already supported as standard categories, so custom categories are unnecessary and add complexity. Option D is wrong because image moderation is a separate feature for analyzing visual content, but the question focuses on text-based hate speech and self-harm detection; enabling it alone wouldn't address the text requirement.

Full explanation →
111
MCQmedium

A developer is building a multilingual chatbot using Azure AI Language. The bot must detect the user's language automatically and route the query to the appropriate language-specific model. Which Azure AI Language feature should the developer use?

A.Translator API.
B.Conversational language understanding (CLU) with multilingual project.
C.Language detection API.
D.Custom text classification model.
AnswerC

Identifies the language of the input text.

Why this answer

Option A is correct because language detection identifies the language of the input text, allowing routing to the correct model. Option B is wrong because translation changes the text but does not detect language. Option C is wrong because the multilingual CLU model handles multiple languages in one project but does not detect language.

Option D is wrong because the custom text classification model is for classification, not language detection.

Full explanation →
112
MCQhard

You are developing a bot using Microsoft Bot Framework and Azure AI Language. The bot must handle user intents that change mid-conversation. Which feature should you implement?

A.Prompt dialogs
B.Waterfall dialogs
C.Adaptive dialogs
D.QnA Maker knowledge base
AnswerC

Adaptive dialogs support dynamic conversation flow and can handle interruptions and changing intents.

Why this answer

Adaptive dialogs are designed for dynamic, event-driven conversations where user intents can change mid-conversation. They use a trigger-based model (e.g., onIntent, onTurn) that allows the bot to react to new intents at any point, unlike linear dialog models. This makes them ideal for handling mid-conversation intent shifts without requiring predefined dialog flows.

Exam trap

The trap here is that candidates often confuse waterfall dialogs (which are sequential and rigid) with adaptive dialogs (which are event-driven and flexible), assuming any dialog can handle mid-conversation changes, but only adaptive dialogs support dynamic interruption and re-routing.

How to eliminate wrong answers

Option A is wrong because prompt dialogs are simple, reusable components for collecting a single piece of input (e.g., text, number) and do not handle intent changes mid-conversation. Option B is wrong because waterfall dialogs follow a fixed, sequential step-by-step flow and cannot dynamically redirect to a different intent once started; they are designed for linear, predictable interactions. Option D is wrong because QnA Maker knowledge base is a question-answering service that matches user queries to predefined Q&A pairs; it does not manage conversational state or handle intent routing.

Full explanation →
113
MCQmedium

You are developing an Azure AI Language solution to analyze customer support tickets. Each ticket has a subject and a description. You need to automatically classify tickets into categories (e.g., 'billing', 'technical', 'account') and extract the product name mentioned. You have a labeled dataset of 10,000 tickets with category labels and product name annotations. The solution must be cost-effective and easy to retrain as new categories emerge. You want to use a single Azure AI Language resource. Which approach should you use?

A.Use custom text classification for category and key phrase extraction for product name.
B.Use conversational language understanding (CLU) to handle both classification and entity extraction in a single model.
C.Use custom text classification for category and custom named entity recognition for product name extraction.
D.Use custom text classification for category and prebuilt named entity recognition for product name extraction.
AnswerC

Both custom text classification and custom NER are available in Azure AI Language and can be trained on the labeled dataset to meet the requirements.

Why this answer

Option C is correct because custom text classification can handle the classification task, and custom NER can extract product names, both within the same Azure AI Language resource. Option A is wrong because the prebuilt NER cannot extract custom product names. Option B is wrong because key phrase extraction may not reliably extract product names.

Option D is wrong because conversational language understanding is designed for conversational flows, not single-turn ticket classification.

Full explanation →
114
MCQeasy

Your knowledge mining pipeline uses Azure AI Search to index PDF files. You need to extract text from the PDFs and also recognize embedded tables. Which built-in skill should you use?

A.Document Extraction skill
B.OCR skill
C.Custom Web API skill
D.Entity Recognition skill
AnswerA

This skill extracts text from files like PDFs; for tables, you need additional processing like Document Intelligence.

Why this answer

The Document Extraction skill extracts text from PDFs and other file types. For tables, you would use Document Intelligence. The other skills are for different purposes.

Full explanation →
115
Multi-Selectmedium

You need to design an Azure AI solution that processes sensitive customer data. The solution must comply with GDPR and data residency requirements. Which TWO actions should you take?

Select 2 answers
A.Enable customer-managed keys (CMK) for the AI resource
B.Use Microsoft Entra ID for authentication
C.Use a private endpoint to connect to the AI resource
D.Deploy the AI resource in the region where the data originates
E.Configure data retention and deletion policies
AnswersD, E

Ensures data residency.

Why this answer

Option D is correct because deploying the AI resource in the region where the data originates ensures compliance with GDPR data residency requirements, which mandate that personal data must not leave the geographic region of the data subject. This directly addresses the principle of data localization, a core GDPR obligation for controllers and processors.

Exam trap

The trap here is that candidates confuse security controls (CMK, private endpoints, Entra ID) with data residency and compliance requirements, assuming any security measure satisfies GDPR, when in fact GDPR specifically mandates geographic data localization and retention policies.

Full explanation →
116
MCQmedium

You notice a spike in errors (HTTP 429) on a specific day. What is the most likely cause?

A.Network connectivity issues.
B.The number of calls exceeded the rate limit for the service tier.
C.Authentication tokens expired.
D.Invalid API keys were used.
AnswerB

HTTP 429 indicates rate limiting.

Why this answer

Option C is correct because HTTP 429 indicates rate limiting, meaning the number of calls exceeded the allowed limit for the service tier. Option A is wrong because invalid API keys would return 401, not 429. Option B is wrong because authentication failures return 401 or 403.

Option D is wrong because network issues cause timeouts or 5xx errors, not 429.

Full explanation →
117
MCQhard

You run the Azure CLI command shown in the exhibit. After a few minutes, the deployment fails with a quota error. What is the most likely cause?

A.The SKU name 'Standard' is invalid for Azure OpenAI deployments.
B.The model version '0613' is deprecated and no longer available.
C.The requested capacity of 10 exceeds the available quota for the gpt-4 model in that region.
D.The resource group name 'myResourceGroup' does not exist.
AnswerC

Quota errors occur when capacity exceeds regional limits.

Why this answer

The quota error indicates that the requested capacity (10 units) for the gpt-4 model exceeds the available quota in the target region. Azure OpenAI deployments require sufficient model-specific quota, which is region- and model-specific. The error is not related to SKU name validity, model version deprecation, or resource group existence.

Exam trap

The trap here is that candidates might confuse a quota error with a model deprecation or SKU issue, but the error message's explicit mention of 'quota' directly points to capacity limits, not configuration or availability problems.

How to eliminate wrong answers

Option A is wrong because 'Standard' is a valid SKU name for Azure OpenAI deployments; the error message specifically mentions quota, not an invalid SKU. Option B is wrong because model version '0613' is a valid and available version for gpt-4; deprecation would produce a different error (e.g., 'ModelNotFound'), not a quota error. Option D is wrong because if the resource group did not exist, the Azure CLI would fail immediately with a 'ResourceGroupNotFound' error, not after several minutes with a quota error.

Full explanation →
118
MCQmedium

Refer to the exhibit. You are reviewing a Bicep template for deploying an Azure AI Language resource. After deployment, you need to ensure that the resource uses a private endpoint to block public access. Which additional resource should you include in the template?

A.A service endpoint for Microsoft.CognitiveServices
B.A virtual network peering connection
C.A virtual network gateway
D.A Private Endpoint resource linked to the Cognitive Services account
AnswerD

Private endpoint enables private access and disables public access.

Why this answer

Option D is correct because a Private Endpoint resource, when linked to the Cognitive Services account via the `privateLinkServiceId` property, assigns a private IP address from a virtual network to the Azure AI Language resource. This blocks all public access by default when the resource's `publicNetworkAccess` property is set to 'Disabled', ensuring traffic only flows over the Microsoft backbone network through Azure Private Link.

Exam trap

The trap here is that candidates confuse service endpoints (which only filter source traffic but leave the public endpoint active) with private endpoints (which completely remove public accessibility), leading them to incorrectly select Option A.

How to eliminate wrong answers

Option A is wrong because a service endpoint for Microsoft.CognitiveServices does not block public access; it only restricts source traffic to a specific virtual network subnet while still allowing public endpoints to be reachable from the internet. Option B is wrong because virtual network peering connects two virtual networks but does not provide a private IP address or block public access to an Azure AI resource. Option C is wrong because a virtual network gateway is used for site-to-site VPN or ExpressRoute connections, not for creating a private endpoint to an Azure PaaS service.

Full explanation →
119
MCQhard

You are designing a knowledge mining solution that must extract entities from scanned handwritten forms. The forms contain signatures and checkboxes. Which combination of Azure AI services should you recommend?

A.Azure AI Document Intelligence with a custom neural model and Azure AI Language for entity linking
B.Azure AI Document Intelligence with a premade model and Azure AI Computer Vision
C.Azure AI Computer Vision (OCR) and Azure AI Search with integrated vectorization
D.Azure Cognitive Search and Azure AI Document Intelligence with a premade model
AnswerA

Custom neural models support handwriting; Language can enrich entities.

Why this answer

Option A is correct because Document Intelligence can extract handwriting and layout, and AI Language can post-process entities. Option B is wrong because Computer Vision OCR is for printed text only. Option C is wrong because Cognitive Search is not an extraction service.

Option D is wrong because AI Document Intelligence already includes OCR; adding Computer Vision is redundant.

Full explanation →
120
MCQeasy

A university is developing an app for students to take photos of handwritten notes and convert them to digital text. The app must support multiple languages including English and Spanish. The solution should use a pre-built AI service. Which Azure service should you use?

A.Azure AI Document Intelligence with a custom model
B.Azure AI Vision Read API (OCR)
C.Azure AI Language with custom entity recognition
D.Custom Vision with a custom handwriting recognition model
AnswerB

Supports handwritten text and multiple languages.

Why this answer

Azure AI Vision OCR API supports multiple languages and handwritten text. Custom Vision does not support OCR. Azure AI Language requires text input.

Azure AI Document Intelligence is for structured documents.

Full explanation →
121
Multi-Selectmedium

Which TWO actions should you take to reduce the latency of an Azure AI Computer Vision OCR call on a large image?

Select 2 answers
A.Use a CPU-bound compute instance.
B.Resize the image to a smaller resolution before calling the API.
C.Increase the API timeout value.
D.Use the Read API asynchronously.
E.Deploy the Cognitive Services container on-premises.
AnswersB, D

Smaller images process faster.

Why this answer

Options B and D are correct. B: Resizing reduces processing time. D: Using the Read API asynchronously allows the client to poll, avoiding timeout.

A: Increasing timeout doesn't reduce latency. C: Using CPU doesn't help. E: Cognitive Services container on-premises might add network latency.

Full explanation →
122
Multi-Selectmedium

Which THREE are valid uses of Azure AI Vision Image Analysis 4.0? (Select three.)

Select 3 answers
A.Extract printed text from an image using OCR
B.Transcribe spoken audio from a video file
C.Detect objects in an image and return bounding boxes
D.Generate a human-readable caption for an image
E.Translate text found in an image to another language
AnswersA, C, D

OCR is included in Image Analysis 4.0.

Why this answer

Options A, B, and D are correct. Image Analysis 4.0 can generate captions, detect objects, and read text. Option C is wrong because translating text is done by Azure AI Translator.

Option E is wrong because transcribing audio is done by Azure AI Speech.

Full explanation →
123
MCQeasy

You are building an application that needs to extract printed text from scanned invoices. Which Azure AI service should you use?

A.Azure AI Vision OCR
B.Azure AI Document Intelligence
C.Azure AI Search
D.Azure AI Language
AnswerA

OCR in Azure AI Vision is designed to extract printed text from images.

Why this answer

Azure AI Vision's OCR reads printed text from images. Azure AI Document Intelligence is for forms and structured documents, but OCR is the core capability for printed text extraction.

Full explanation →
124
Multi-Selectmedium

Which TWO actions are required to enable private endpoint connectivity for an Azure AI Language resource?

Select 2 answers
A.Disable public network access on the AI resource
B.Create the private endpoint in a subnet of a virtual network
C.Configure an Azure Firewall rule to allow the private endpoint
D.Create a private DNS zone for the resource
E.Add a service tag to the network security group
AnswersA, B

To enforce private endpoint, public access must be disabled.

Why this answer

Private endpoint requires a subnet in a VNet and disabling public network access. Option A is wrong because private DNS zone is optional but recommended. Option D is wrong because service tags are not required.

Option E is wrong because firewall rules are not needed.

Full explanation →
125
MCQmedium

You have an Azure AI Search indexer that is configured to index PDF files from Azure Blob Storage. The indexer is not extracting any text from the PDFs, and no errors are reported. You review the indexer definition as shown. What is the most likely cause?

A.The parsingMode is set to 'json' instead of 'default' or 'text'
B.The field mapping from 'content' to 'content' is redundant and causes a conflict
C.The field mapping for 'metadata_storage_path' should be to 'metadata_storage_path'
D.The dataToExtract is set to 'contentAndMetadata' which is not supported for PDFs
AnswerA

JSON mode expects JSON files, not PDFs.

Why this answer

Option B is correct because the parsingMode is set to 'json', which is for JSON files, not PDFs. For PDF parsing, the mode should be 'default' or 'text'. Option A is wrong because contentAndMetadata is fine.

Option C is wrong because field mappings are correct. Option D is wrong because the path mapping is fine.

Full explanation →
126
MCQeasy

You are developing a generative AI application that must comply with responsible AI principles. Which Azure AI service should you use to detect and filter harmful content in both input prompts and output responses?

A.Microsoft Purview
B.Azure AI Content Safety
C.Azure OpenAI Service
D.Azure AI Language
AnswerB

Content Safety is designed to detect and filter harmful content.

Why this answer

Option D is correct because Azure AI Content Safety provides content filtering for harmful categories. Option A is wrong because Azure OpenAI Service provides the model, not content safety. Option B is wrong because Azure AI Language offers moderation APIs but Content Safety is dedicated.

Option C is wrong because Microsoft Purview is for data governance, not content safety.

Full explanation →
127
MCQmedium

Refer to the exhibit. You deploy this ARM template to create an agent in Microsoft Foundry. The agent uses an Azure Function as an action. The deployment succeeds but when the agent calls the function, it gets a 401 Unauthorized error. What is the most likely cause?

A.The authentication type should be ManagedIdentity instead of ApiKey.
B.The model provider is AzureOpenAI but should be AzureAI.
C.The URL parameter is missing the function name in the path.
D.The listKeys function returns the host key, but the function requires a function-level key.
AnswerD

The function may be configured with function-level authorization requiring a specific key.

Why this answer

Option A is correct because the function key is retrieved from the default key, but the function might require a specific key. Option B is wrong because the function key is correctly passed as a header. Option C is wrong because the endpoint is valid.

Option D is wrong because the model name is valid.

Full explanation →
128
MCQeasy

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

A.Overfitting to the training data
B.Data leakage between training and test sets
C.Insufficient training data
D.Label imbalance in the training data
AnswerA

Overfitting causes high accuracy on training/test sets but poor generalization.

Why this answer

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

Exam trap

The trap here is that candidates often confuse overfitting with data leakage or label imbalance, but the key diagnostic is the large gap between high test accuracy and low validation accuracy, which uniquely points to overfitting.

How to eliminate wrong answers

Option B is wrong because data leakage between training and test sets would inflate both test and validation accuracy, not create a large gap between them. Option C is wrong because insufficient training data typically causes underfitting (low accuracy on both test and validation sets), not a high test accuracy with low validation accuracy. Option D is wrong because label imbalance in the training data would cause the model to be biased toward the majority class, leading to poor performance on minority classes across both test and validation sets, not a discrepancy between them.

Full explanation →
129
Multi-Selectmedium

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

Select 2 answers
A.Use provisioned throughput with reserved capacity.
B.Increase the temperature parameter.
C.Use a smaller model like GPT-3.5 instead of GPT-4.
D.Increase the max_tokens parameter to get longer responses.
E.Enable content filtering on all requests.
AnswersA, C

Reserved capacity offers a discount compared to pay-as-you-go.

Why this answer

Options A and D are correct. A: Using a smaller model like GPT-3.5 instead of GPT-4 reduces cost per token. D: Provisioned throughput with reserved capacity offers lower per-token cost for high usage.

B is wrong because increasing max_tokens increases cost. C is wrong because using a higher temperature does not affect token count. E is wrong because content filters do not reduce token consumption.

Full explanation →
130
MCQhard

You are building a custom text classification solution in Azure AI Language. You have a dataset with 10 categories and 1000 labeled documents. You need to choose the best project type. What should you use?

A.Conversational Language Understanding (CLU)
B.Key Phrase Extraction
C.Prebuilt Text Classification API
D.Custom text classification (single or multi-label)
AnswerD

Custom text classification can be trained on your own categories and labels.

Why this answer

Option A is correct because Custom text classification is designed for custom categories and can handle multiple labels per document if needed. Option B is wrong because the built-in Text Classification API only supports predefined categories. Option C is wrong because Conversational Language Understanding is for intents and entities, not text classification.

Option D is wrong because Key Phrase Extraction is not a classification feature.

Full explanation →
131
MCQhard

You are deploying a conversational AI solution using Microsoft Copilot Studio. The solution must comply with organizational data loss prevention (DLP) policies by preventing sensitive data from being sent to the underlying Azure OpenAI model. What should you configure?

A.Configure content filters in Azure OpenAI Studio
B.Define DLP policies in Microsoft 365 compliance center and apply to Copilot Studio
C.Enable Azure AI Content Safety in the bot's generative AI configuration
D.Set the temperature parameter to 0 to reduce variability
AnswerB

DLP policies in M365 can block sensitive data from being sent to AI models.

Why this answer

Option B is correct because Microsoft Copilot Studio integrates with Microsoft 365 DLP policies to prevent sensitive data from being sent to the underlying Azure OpenAI model. By defining DLP policies in the Microsoft 365 compliance center and applying them to Copilot Studio, you can enforce data loss prevention rules that block or restrict the transmission of sensitive information (e.g., credit card numbers, social security numbers) to the generative AI backend. This ensures compliance with organizational security requirements without modifying the AI model itself.

Exam trap

The trap here is that candidates confuse Azure AI Content Safety (which handles harmful content moderation) with DLP policies (which handle sensitive data protection), leading them to select Option C instead of the correct DLP-based approach.

How to eliminate wrong answers

Option A is wrong because content filters in Azure OpenAI Studio are designed to filter harmful or offensive content in model outputs, not to prevent sensitive data from being sent to the model as input; they operate on the response side, not the request side. Option C is wrong because Azure AI Content Safety is a service for detecting and filtering harmful content (e.g., hate speech, violence) in both inputs and outputs, but it does not enforce DLP policies or block sensitive data based on organizational compliance rules; it focuses on safety, not data loss prevention. Option D is wrong because setting the temperature parameter to 0 reduces the randomness of the model's responses, making them more deterministic, but it has no effect on preventing sensitive data from being sent to the model; it controls output variability, not input filtering.

Full explanation →
132
Multi-Selecteasy

Which TWO components are required to create a custom text classification model in Azure AI Language?

Select 2 answers
A.A set of labeled documents
B.A QnA Maker knowledge base
C.A project in Azure AI Language
D.A Language Understanding (LUIS) app
E.An Azure Functions app
AnswersA, C

Labeled documents are required for training.

Why this answer

A custom text classification project requires a project in Azure AI Language and a set of labeled documents for training. A LUIS app and a QnA maker knowledge base are different services. An Azure Functions app is not required.

Full explanation →
133
Multi-Selectmedium

Which TWO Azure AI services can be used together to build a knowledge mining solution that extracts text from handwritten notes and indexes them for search?

Select 2 answers
A.Azure AI Search
B.Azure AI Document Intelligence
C.Azure AI Vision (Read API)
D.Azure AI Language
E.Azure AI Translator
AnswersA, C

Indexes extracted text.

Why this answer

Azure AI Vision Read API can extract text from handwritten notes (OCR). Azure AI Search indexes the extracted text. Option A is incorrect because Document Intelligence is for forms, not handwriting.

Option C is incorrect because Language is for text analysis. Option D is incorrect because Translator is for translation.

Full explanation →
134
MCQhard

Your Azure AI Language custom entity recognition model incorrectly extracts 'Microsoft' as an organization when it refers to the company, but fails to extract 'Microsoft' as a product when it refers to the software. How should you improve the model?

A.Reduce the amount of training data to avoid confusion
B.Remove the 'Organization' entity type from the model
C.Add more training sentences without labeling the entity type
D.Label 'Microsoft' as both 'Organization' and 'Product' in different training sentences with appropriate context
AnswerD

Providing multiple entity types for the same word helps the model learn context-based disambiguation.

Why this answer

Option D is correct because custom entity recognition models in Azure AI Language learn to distinguish entity types based on context. By labeling 'Microsoft' as 'Organization' in sentences where it refers to the company and as 'Product' in sentences where it refers to the software, you provide the model with the contextual clues needed to disambiguate the same token across different uses. This supervised learning approach directly addresses the model's failure to recognize the product entity.

Exam trap

The trap here is that candidates may think reducing data or removing entity types simplifies the problem, but Azure AI Language models require diverse, labeled examples with context to handle polysemy (same word, different meanings).

How to eliminate wrong answers

Option A is wrong because reducing training data would likely worsen model performance by removing valuable examples, not resolve the ambiguity. Option B is wrong because removing the 'Organization' entity type would prevent the model from correctly extracting 'Microsoft' as an organization, which is a valid extraction in many contexts, and does not solve the product extraction issue. Option C is wrong because adding training sentences without labeling the entity type provides no supervised signal for the model to learn the distinction between 'Organization' and 'Product' for the same token.

Full explanation →
135
MCQhard

You are a developer at a global e-commerce company. You are building a multilingual chatbot using Azure AI Language that supports English, French, German, and Spanish. The chatbot must answer frequently asked questions about order status, returns, and shipping. You plan to use Custom Question Answering with a single project containing questions and answers in all four languages. However, during testing, you notice that queries in French and German often return incorrect answers or no answer, while English and Spanish work well. You need to ensure accurate answers across all four languages. What should you do?

A.Create a separate Custom Question Answering project for each language and route user queries to the appropriate project based on language detection.
B.Use Azure Cognitive Search with semantic ranking to index the QnA pairs.
C.Add synonyms in the project for French and German terms to improve matching.
D.Deploy the same project to multiple regions and use traffic manager.
AnswerA

Separate projects ensure optimal language-specific models and accurate answers.

Why this answer

Custom Question Answering is language-specific. Each project is optimized for a single language. Mixing multiple languages in one project degrades performance because the model's language detection and matching algorithms are tuned for one language.

The correct approach is to create separate projects per language and route queries based on detected language. Option A is correct. Option B is incorrect because adding synonyms does not solve the fundamental language mismatch.

Option C is incorrect because the same data will still be mixed. Option D is incorrect because Cognitive Search does not provide the same QnA matching capabilities.

Full explanation →
136
Multi-Selecthard

Which THREE considerations are essential when designing a cost management strategy for an enterprise Azure AI solution that uses multiple AI services, including Azure OpenAI Service, Azure AI Language, and Azure AI Vision?

Select 3 answers
A.Use the Free tier for all services to minimize upfront costs.
B.Choose provisioned throughput units (PTUs) for Azure OpenAI Service for predictable workloads.
C.Use a single multi-service Cognitive Services account to consolidate usage.
D.Deploy Azure Site Recovery to replicate AI services across regions.
E.Enable autoscaling on Azure AI Language to adjust capacity based on demand.
AnswersB, C, E

PTUs provide cost savings over pay-as-you-go for steady usage.

Why this answer

Option B is correct because provisioned throughput units (PTUs) for Azure OpenAI Service provide reserved capacity, ensuring predictable performance and cost for stable workloads. PTUs are ideal for enterprise scenarios where latency and throughput must be consistent, as they allocate dedicated model processing capacity and avoid pay-per-token variability.

Exam trap

The trap here is that candidates often confuse the Free tier's suitability for production or assume Azure Site Recovery can replicate stateless API endpoints, when in fact it is designed for infrastructure-level failover, not cost management.

Full explanation →
137
MCQmedium

You are designing a solution that reads handwritten notes from patient intake forms. The solution must handle various handwriting styles. Which Azure AI capability should you use?

A.Azure AI Document Intelligence Read model
B.Azure AI Custom Vision
C.Azure AI Vision OCR
D.Azure AI Language
AnswerA

The Read model handles handwriting and printed text.

Why this answer

Azure AI Document Intelligence's Read model supports handwriting recognition. The prebuilt-read model is designed to extract both printed and handwritten text.

Full explanation →
138
Multi-Selecthard

You are designing a generative AI solution using Azure OpenAI Service. The solution must support multiple languages and provide consistent quality across languages. Which THREE actions should you take?

Select 3 answers
A.Fine-tune the model on a dataset of a single language
B.Use a model that supports multiple languages (e.g., GPT-4)
C.Provide examples in multiple languages in the prompt
D.Set the temperature to 0 for all requests
E.Test the solution with representative prompts in each language
AnswersB, C, E

Multilingual models handle multiple languages natively.

Why this answer

Option B is correct because GPT-4 is a multilingual model pre-trained on diverse language corpora, enabling it to generate coherent and contextually appropriate responses across many languages without additional fine-tuning. This ensures consistent quality by leveraging the model's inherent cross-lingual capabilities, which is essential for a generative AI solution that must support multiple languages.

Exam trap

The trap here is that candidates may think fine-tuning on a single language (Option A) is sufficient for multilingual support, or that setting temperature to 0 (Option D) universally improves consistency, when in fact these actions undermine the required cross-lingual quality and flexibility.

Full explanation →
139
MCQeasy

Your organization is deploying an agentic solution using Microsoft Copilot Studio. The agent must be able to escalate to a human agent when it cannot resolve a user's request. You need to ensure that the escalation includes the full conversation history. What should you configure?

A.Add an 'End conversation' action and configure a fallback
B.Add a 'Create a ticket' action in the topic
C.Add a 'Start a new topic' action with context variables
D.Add a 'Transfer conversation' action and set it to include the full transcript
AnswerD

Transfer conversation sends history to human agent.

Why this answer

Option C is correct because 'Transfer conversation' automatically passes the full transcript to the human agent. Option A is incorrect because 'Create a ticket' does not transfer the conversation. Option B is incorrect because 'End conversation' terminates without escalation.

Option D is incorrect because 'Start a new topic' does not include history.

Full explanation →
140
MCQeasy

You need to analyze customer call transcripts to identify positive and negative sentiment. Which Azure AI Language feature should you use?

A.Language Detection
B.Named Entity Recognition
C.Key Phrase Extraction
D.Sentiment Analysis
AnswerD

Sentiment Analysis detects positive/negative sentiment.

Why this answer

Option A is correct because Sentiment Analysis is designed for that purpose. Option B is incorrect because Key Phrase Extraction extracts key phrases. Option C is incorrect because Entity Recognition identifies entities.

Option D is incorrect because Language Detection identifies language.

Full explanation →
141
MCQeasy

A company uses Azure AI Search to index customer support tickets. They need to automatically extract key phrases from each ticket to improve search relevance. Which built-in skill should they add to the skillset?

A.Key Phrase Extraction
B.Entity Recognition
C.Sentiment Analysis
D.OCR
AnswerA

Key Phrase Extraction skill extracts key phrases from text.

Why this answer

Option C is correct because the Key Phrase Extraction skill extracts key phrases from text. Option A is incorrect because Entity Recognition extracts named entities. Option B is incorrect because Sentiment Analysis determines sentiment.

Option D is incorrect because OCR extracts text from images.

Full explanation →
142
MCQeasy

You need to generate a summary of a long article using Azure OpenAI. The article is 10,000 tokens long. What should you do to fit the article within the model's context window?

A.Split the article into smaller sections and summarize each section separately.
B.Increase the temperature parameter.
C.Use a model with a smaller context window.
D.Set max_tokens to a lower value.
AnswerA

Chunking the input fits within the context window.

Why this answer

Option A is correct because the article exceeds the model's context window (typically 4096 or 8192 tokens for GPT-3.5/4). Splitting the article into smaller sections and summarizing each separately allows you to process the entire content within the token limits, then combine the summaries for a final coherent output. This is a standard chunking strategy for long documents when using Azure OpenAI.

Exam trap

The trap here is that candidates confuse parameters that control output behavior (temperature, max_tokens) with the fundamental input token limit, leading them to incorrectly believe adjusting these parameters can bypass the context window restriction.

How to eliminate wrong answers

Option B is wrong because increasing the temperature parameter affects randomness and creativity of the output, not the input token limit; it does not help fit a long article into the context window. Option C is wrong because using a model with a smaller context window would make the problem worse, as it reduces the maximum input length, not increase it. Option D is wrong because setting max_tokens to a lower value only truncates the output length, not the input; the article still exceeds the context window and will be rejected or truncated at the input stage.

Full explanation →
143
MCQhard

Your company is building a knowledge base for customer support using Azure AI Search. You have a large dataset of customer emails stored in Azure Blob Storage. The solution must extract key phrases, detect sentiment, and identify customer intents (e.g., complaint, inquiry, feedback). You plan to use built-in AI skills for key phrase extraction and sentiment detection. For intent identification, you need a custom solution because the intents are specific to your business. You have trained a custom Language Understanding (LUIS) model and published it. How should you integrate the LUIS model into the Azure AI Search enrichment pipeline to extract intents?

A.Add a Document Intelligence skill to classify intents.
B.Configure the index to use a custom analyzer to parse intents.
C.Use the built-in Entity Recognition skill to extract intents.
D.Create a custom skill in the skillset that calls the LUIS endpoint and returns the top intent.
AnswerD

Custom skills enable integration with external services like LUIS for custom entity or intent extraction.

Why this answer

Option A is correct because Azure AI Search supports custom skills that can call external APIs, such as a LUIS endpoint. This allows you to integrate the LUIS model for intent extraction. Option B is wrong because LUIS is not a built-in skill.

Option C is wrong because modifying the index does not run enrichment. Option D is wrong because Document Intelligence is not for intent extraction.

Full explanation →
144
Multi-Selectmedium

You are developing an agentic solution that uses multiple agents to handle customer inquiries. You need to ensure that agents can hand off to each other with full context. Which THREE features should you use?

Select 3 answers
A.KQL
B.Shared Memory
C.Agent Handoff
D.Threads
E.Function calling
AnswersB, C, D

Shared Memory allows agents to persist and share state across handoffs.

Why this answer

Agent Handoff, Threads, and Shared Memory are key features for context preservation and handoff. Function calling is for tool integration, not handoff. KQL is for queries.

Full explanation →
145
MCQhard

You have the above Azure AI Search indexer definition. The indexer runs successfully but the 'content' field in the index is empty for all documents. What is the likely cause?

A.The outputFieldMapping references '/document/content' which is not produced by the skillset.
B.The indexer schedule is too frequent.
C.The fieldMapping for 'metadata_storage_path' is incorrect.
D.The batchSize is too large, causing some items to fail silently.
AnswerA

Skillset must output that path.

Why this answer

The outputFieldMapping maps '/document/content' to 'content'. However, the skillset (not shown) might not produce this output, or the skill output path is different. But the exhibit shows that the indexer expects '/document/content' to be available from the skillset.

If the skillset does not output '/document/content', the field remains empty. Option C is correct: the outputFieldMapping references a path that does not exist in the enrichment tree. Option A is incorrect because the field mapping for 'path' is fine.

Option B is incorrect because the indexer runs successfully. Option D is incorrect because batch size does not affect field mapping.

Full explanation →
146
MCQeasy

You are building a chatbot that uses Azure AI Language to extract intents and entities from user utterances. The bot must recognize custom entities like product names that are not in the default model. Which feature should you use?

A.Prebuilt entity recognition component.
B.Key phrase extraction.
C.Custom named entity recognition (NER) component.
D.List entity in a conversational language understanding (CLU) project.
AnswerC

Allows training a model to extract custom entities.

Why this answer

Option C is correct because a custom named entity recognition (NER) component allows you to train a model to extract custom entities. Option A is wrong because the prebuilt entity recognition only recognizes common entities. Option B is wrong because a list entity in a conversational language understanding (CLU) project is for matching exact terms, not for training a model.

Option D is wrong because key phrase extraction does not extract entities.

Full explanation →
147
MCQmedium

Your team uses Azure AI Foundry to deploy a custom chat model. The model must meet compliance by explaining its reasoning and citing sources. Which feature should you enable?

A.Fine-tuning with domain-specific data
B.Content filtering
C.Groundedness detection with citation
D.Prompt engineering
AnswerC

Groundedness detection ensures the model cites sources for compliance.

Why this answer

Groundedness detection with citation is the correct feature because it directly addresses the compliance requirement for the model to explain its reasoning and cite sources. This feature, available in Azure AI Foundry, evaluates the model's responses against the provided grounding documents and automatically generates citations, ensuring that the output is factually supported and traceable.

Exam trap

The trap here is that candidates may confuse content filtering (which blocks unsafe content) with groundedness detection (which ensures factual accuracy and source attribution), leading them to choose option B when the question specifically asks for reasoning and citation capabilities.

How to eliminate wrong answers

Option A is wrong because fine-tuning with domain-specific data improves the model's performance on specialized tasks but does not inherently enforce citation or reasoning transparency; it only adapts the model to a specific dataset. Option B is wrong because content filtering is designed to block harmful or inappropriate content based on predefined categories, not to provide reasoning or source citations for the model's outputs. Option D is wrong because prompt engineering involves crafting input prompts to guide model behavior, but it does not automatically generate citations or ensure that the model's reasoning is explained in a compliant manner; it relies on manual design and does not enforce source attribution.

Full explanation →
148
MCQmedium

You are a solution architect at a news agency. The agency publishes thousands of articles daily. You need to build a knowledge mining solution that enables journalists to search for articles by topic, sentiment, key people, and locations mentioned. The articles are stored as HTML files in Azure Blob Storage. The solution must also provide a summary for each article. You plan to use Azure AI Search with cognitive skills and Azure OpenAI. Which combination of skills and features should you include to meet all requirements with the best performance and accuracy?

A.Use Azure AI Document Intelligence to extract content from HTML, then use Azure AI Language to extract entities and sentiment. Index in Azure AI Search with semantic search.
B.Skillset with Entity Recognition skill, Sentiment skill, Key Phrase Extraction skill, and Text Translation skill. Enable semantic search.
C.Skillset with Entity Recognition skill, Sentiment skill, and Key Phrase Extraction skill. Use Azure OpenAI service to generate summaries via a custom skill that calls the GPT model. Enable semantic search.
D.Skillset with Entity Recognition skill, Sentiment skill, and Text Analytics for Health skill to extract medical terms. Use Azure OpenAI for summarization as a custom skill.
AnswerC

Covers all requirements: topics, sentiment, entities, and summarization.

Why this answer

Option C is correct because it uses Entity Recognition for people/locations, Sentiment for sentiment, Key Phrase Extraction for topics, and Azure OpenAI for summarization (via a custom skill). Option A uses Translator unnecessarily. Option B uses Text Analytics for health (not appropriate) and lacks summarization.

Option D uses Document Intelligence which is for documents, not HTML.

Full explanation →
149
Multi-Selecteasy

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

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

Document Intelligence extracts text from documents.

Why this answer

Azure AI Document Intelligence (formerly Form Recognizer) includes the Read OCR engine that extracts printed and handwritten text from images and documents. Azure AI Computer Vision provides the OCR API (optical character recognition) which can extract text from images, including both printed and handwritten text, and supports multiple languages. Both services are designed specifically for text extraction from visual content.

Exam trap

The trap here is that candidates may confuse Azure AI Video Indexer's ability to extract text from video frames as a primary image text extraction service, but it is designed for video analysis and indexing, not standalone image text extraction.

Full explanation →
150
MCQhard

A legal firm uses Azure AI Language's custom NER to extract party names, dates, and clauses from contracts. The model performs well on English contracts but poorly on French contracts. The firm wants to improve performance without retraining from scratch. What is the most efficient approach?

A.Create a separate custom NER project for French and train from scratch using French contracts.
B.Retrain the English model with a mix of English and French contracts.
C.Use Azure AI Translator to translate French contracts to English, then use the English model.
D.Use the multilingual option in Azure AI Language custom NER to extend the existing project to include French.
AnswerD

Multilingual projects allow extending to other languages leveraging existing training.

Why this answer

Option D is correct because Azure AI Language supports multilingual projects that can learn from pre-existing English data and transfer to French. Option A is wrong because it requires separate management. Option B is wrong because translation may lose legal nuances.

Option C is wrong because Translator adds latency and cost.

Full explanation →
Page 1

Page 2 of 14

Page 3

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 →