Scenario practice questions

A developer is building a mobile app that uses Azure Computer Vision to analyze images. The app needs to handle many requests with low latency. Which pricing tier should they choose?

Trap 1: Video Analyzer S1 tier.

For video indexing, not image analysis.

Trap 2: Free F0 tier.

Free tier has 20 calls per minute limit.

Trap 3: Custom Vision S0 tier.

Custom Vision is separate from Computer Vision.

A
Video Analyzer S1 tier.
Why wrong: For video indexing, not image analysis.
B
Free F0 tier.
Why wrong: Free tier has 20 calls per minute limit.
C
Custom Vision S0 tier.
Why wrong: Custom Vision is separate from Computer Vision.
D
Computer Vision S1 tier.
S1 offers up to 30 calls per second.

Question 2mediummultiple choice

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?

Trap 1: Azure AI Speech

Speech service is for audio processing, not text translation.

Trap 2: Azure OpenAI Service

OpenAI can translate but is not the dedicated service; Translator is optimized for this.

Trap 3: Azure AI Language

Language service does not include translator; it provides NLP features like sentiment analysis.

A
Azure AI Speech
Why wrong: Speech service is for audio processing, not text translation.
B
Azure OpenAI Service
Why wrong: OpenAI can translate but is not the dedicated service; Translator is optimized for this.
C
Azure AI Language
Why wrong: Language service does not include translator; it provides NLP features like sentiment analysis.
D
Azure AI Translator
Translator resource is designed for real-time text translation.

Question 3hardmulti select

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?

Trap 1: Use the Free tier for all services to minimize upfront costs.

Free tier has strict limits and is not scalable for enterprise.

Trap 2: Deploy Azure Site Recovery to replicate AI services across regions.

Site Recovery is for disaster recovery, not cost optimization.

A
Use the Free tier for all services to minimize upfront costs.
Why wrong: Free tier has strict limits and is not scalable for enterprise.
B
Choose provisioned throughput units (PTUs) for Azure OpenAI Service for predictable workloads.
PTUs provide cost savings over pay-as-you-go for steady usage.
C
Use a single multi-service Cognitive Services account to consolidate usage.
Multi-service account simplifies management and may offer tiered pricing.
D
Deploy Azure Site Recovery to replicate AI services across regions.
Why wrong: Site Recovery is for disaster recovery, not cost optimization.
E
Enable autoscaling on Azure AI Language to adjust capacity based on demand.
Autoscaling prevents paying for unused capacity.

Question 4mediummultiple choice

Your team is developing a chatbot using Azure AI Bot Service with language understanding via Azure AI Language (CLU). You need to ensure that the chatbot can handle multiple intents in a single user utterance and return confidence scores for each. Which configuration should you use?

Trap 1: Use an Orchestration workflow project that connects to multiple CLU…

Orchestration routes to different projects but does not return multiple intents from a single utterance.

Trap 2: Use a standard CLU project and manually combine intents.

Standard CLU returns a single predicted intent per utterance.

Trap 3: Use Azure AI QnA Maker with custom logic to detect intents.

QnA Maker does not support intent detection.

A
Use a Conversational Language Understanding (CLU) project with the 'Multiple intents' setting enabled.
CLU supports multiple intents per utterance with confidence scores.
B
Use an Orchestration workflow project that connects to multiple CLU projects.
Why wrong: Orchestration routes to different projects but does not return multiple intents from a single utterance.
C
Use a standard CLU project and manually combine intents.
Why wrong: Standard CLU returns a single predicted intent per utterance.
D
Use Azure AI QnA Maker with custom logic to detect intents.
Why wrong: QnA Maker does not support intent detection.

Question 5mediummultiple choice

You are deploying a conversational AI chatbot using Azure AI Language service. The chatbot must be able to switch between multiple intents in a single conversation without restarting the session. Which feature should you enable?

Trap 1: Orchestration workflow

Orchestration routes between different skills, not multi-intent within one skill.

Trap 2: Dynamic entity extraction

Entity extraction identifies entities, not intents.

Trap 3: Prebuilt domain components

Prebuilt domains are static and don't handle multi-intent transitions.

A
Active learning
Active learning enables the model to learn from conversational context and handle multiple intents.
B
Orchestration workflow
Why wrong: Orchestration routes between different skills, not multi-intent within one skill.
C
Dynamic entity extraction
Why wrong: Entity extraction identifies entities, not intents.
D
Prebuilt domain components
Why wrong: Prebuilt domains are static and don't handle multi-intent transitions.

Question 6hardmultiple choice

You are developing a custom agent using the Microsoft Bot Framework SDK. The agent must handle multiple languages and use the Azure AI Translator service to translate user messages to English before processing. The solution should minimize latency. Where should the translation logic be implemented?

Trap 1: In the OnMessageActivityAsync method after receiving the user's…

This would work but is not the most efficient; middleware is designed for cross-cutting concerns.

Trap 2: By using the Bot Framework's built-in language detection feature.

Language detection does not translate; it only detects the language.

Trap 3: By sending the user's message to a separate translation endpoint…

Client-side translation adds complexity and may not be feasible if the client is a channel like Teams.

A
As a custom middleware component that intercepts incoming activities and translates them.
Middleware runs before the bot logic, translating messages efficiently without affecting the rest of the code.
B
In the OnMessageActivityAsync method after receiving the user's message.
Why wrong: This would work but is not the most efficient; middleware is designed for cross-cutting concerns.
C
By using the Bot Framework's built-in language detection feature.
Why wrong: Language detection does not translate; it only detects the language.
D
By sending the user's message to a separate translation endpoint from the client application.
Why wrong: Client-side translation adds complexity and may not be feasible if the client is a channel like Teams.

Question 7mediummultiple choice

An organization uses Microsoft Copilot Studio to create an agent for IT support. The agent should be able to reset passwords, unlock accounts, and look up user information by connecting to on-premises Active Directory via Microsoft Entra ID. Which type of authentication should be used for the agent to access these actions?

Trap 1: Basic authentication with username and password.

Basic authentication is insecure and not recommended.

Trap 2: API key authentication to the on-premises API.

API keys do not support user-specific authorization and are not secure for this scenario.

Trap 3: Certificate-based authentication for the agent's identity.

Certificate-based auth is for server-to-server, not for user delegation.

A
Basic authentication with username and password.
Why wrong: Basic authentication is insecure and not recommended.
B
API key authentication to the on-premises API.
Why wrong: API keys do not support user-specific authorization and are not secure for this scenario.
C
Certificate-based authentication for the agent's identity.
Why wrong: Certificate-based auth is for server-to-server, not for user delegation.
D
OAuth 2.0 authentication with Microsoft Entra ID using on-behalf-of flow.
On-behalf-of flow allows the agent to obtain tokens to call downstream APIs on behalf of the user.

Read the full NAT/PAT explanation →

Question 8mediummultiple choice

You have a computer vision solution that analyzes security camera feeds to detect people and vehicles. The solution uses Azure AI Vision Spatial Analysis. You need to ensure compliance with privacy regulations by blurring detected faces. Which feature should you enable?

Trap 1: Use Azure AI Content Safety to filter faces

Content Safety is for harmful content, not face blurring.

Trap 2: Post-process frames with Azure AI Face client SDK

Post-processing adds latency; built-in blurring is more efficient.

Trap 3: Enable face detection and redact faces using Azure AI Video Indexer

Video Indexer is for media analysis, not real-time spatial analysis.

A
Use Azure AI Content Safety to filter faces
Why wrong: Content Safety is for harmful content, not face blurring.
B
Post-process frames with Azure AI Face client SDK
Why wrong: Post-processing adds latency; built-in blurring is more efficient.
C
Enable face detection and redact faces using Azure AI Video Indexer
Why wrong: Video Indexer is for media analysis, not real-time spatial analysis.
D
Enable face blurring in the Spatial Analysis configuration
Spatial Analysis supports face blurring to obscure identities.

Question 9mediummultiple choice

You are building a solution to automatically tag images uploaded to an Azure Storage blob container using Azure AI Vision. The solution must process images as soon as they are uploaded. Which service should you use to trigger the image analysis?

Trap 1: Azure Functions with a timer trigger

Timer trigger is scheduled.

Trap 2: Azure Batch with a job schedule

Batch is for batch processing.

Trap 3: Azure Logic Apps with a recurrence trigger

Recurrence trigger is scheduled, not event-driven.

A
Azure Functions with a timer trigger
Why wrong: Timer trigger is scheduled.
B
Azure Event Grid with an Azure Function trigger
Event Grid provides real-time event-driven trigger.
C
Azure Batch with a job schedule
Why wrong: Batch is for batch processing.
D
Azure Logic Apps with a recurrence trigger
Why wrong: Recurrence trigger is scheduled, not event-driven.

Question 10mediummultiple choice

You are deploying a question answering solution using Azure AI Language. The solution must be able to provide answers from a set of frequently asked questions (FAQs) in PDF format. What should you do?

Trap 1: Use Azure AI Search with cognitive skills.

Azure AI Search is for search, not question answering.

Trap 2: Use Azure OpenAI with a system prompt containing the PDFs.

Requires prompt engineering and may not use PDFs directly.

Trap 3: Use the pre-built question answering in Azure AI Language.

Pre-built is for common scenarios, not custom PDFs.

A
Use Azure AI Search with cognitive skills.
Why wrong: Azure AI Search is for search, not question answering.
B
Create a custom question answering project and add the PDF as a source.
Custom question answering can ingest PDFs.
C
Use Azure OpenAI with a system prompt containing the PDFs.
Why wrong: Requires prompt engineering and may not use PDFs directly.
D
Use the pre-built question answering in Azure AI Language.
Why wrong: Pre-built is for common scenarios, not custom PDFs.

Question 11mediummultiple choice

You are building a knowledge mining solution to extract insights from a large set of PDF contracts. The solution must identify parties, dates, and monetary amounts. Which Azure AI service should you use as the primary extraction engine?

Trap 1: Azure AI Language (custom NER)

Custom NER is text-based and misses layout features like tables.

Trap 2: Azure AI Search with integrated vectorization

Azure AI Search is for indexing and search, not primary extraction.

Trap 3: Azure OpenAI Service with GPT-4o

Overkill and higher cost; Document Intelligence is purpose-built.

A
Azure AI Language (custom NER)
Why wrong: Custom NER is text-based and misses layout features like tables.
B
Azure AI Search with integrated vectorization
Why wrong: Azure AI Search is for indexing and search, not primary extraction.
C
Azure OpenAI Service with GPT-4o
Why wrong: Overkill and higher cost; Document Intelligence is purpose-built.
D
Azure AI Document Intelligence
Designed for extracting fields from forms and documents.