Microsoft Developing AI Cloud Solutions on Azure AI-200 Exam Questions

Page: 1 / 14
Total 142 questions
Question 1

An application performs similarity search across 5 million embeddings stored in Azure Database for PostgreSQL with pgvector. Queries often filter by department before ranking by cosine distance.

P95 latency for vector similarity queries exceeds the SLA target. Monitoring shows sustained high CPU use during query execution.

You need to reduce P95 latency for filtered vector similarity queries.

What should you do?



Answer : C


Question 2

You are designing an Azure Function that will process orders from a new Azure Service Bus queue.

You need to prevent the system from processing messages more than once and preserve failed messages for investigation.

Which two actions should you implement? Each correct answer presents part of the solution.

NOTE: Each correct selection is worth one point.



Answer : A, C

Enable duplicate detection on the Azure Service Bus queue to suppress duplicate sends. Service Bus tracks application-defined MessageId values during the configured duplicate-detection window. If another message with the same MessageId arrives during that interval, Service Bus accepts the send operation but discards the duplicate message. Microsoft specifically recommends using a deterministic business identifier, such as an order ID, as the basis for MessageId when duplicate suppression is required. Duplicate detection must be enabled when the queue is created and is supported in the Standard and Premium tiers.

For failed or poison messages, implement dead-letter handling. A Service Bus queue has a dead-letter subqueue that retains messages that cannot be successfully processed, including messages that exceed the configured maximum delivery count. Those messages remain available for inspection, troubleshooting, correction, or controlled reprocessing rather than being silently discarded.

Sessions provide ordered/FIFO processing and session affinity; they do not primarily prevent duplicates. An HTTP trigger would bypass the required Service Bus processing architecture.

For strict production-grade exactly-once business effects, the consumer should additionally be idempotent, because a message can still be redelivered after a lock or settlement failure.

Study Guide references: Azure Service Bus duplicate detection; MessageId; dead-letter queues; poison-message handling; reliable asynchronous processing.


Question 3

You are developing an application that must extract structured field data (invoice number, total, vendor name) from scanned invoices in multiple layouts.

You need a solution that requires no custom model training.

What should you use?



Answer : B

Use the Azure AI Document Intelligence prebuilt invoice model. The prebuilt invoice model is specifically designed to analyze invoices and extract structured fields and line items without requiring an organization to train a custom model. Microsoft documents support for fields including InvoiceId, VendorName, InvoiceTotal, InvoiceDate, CustomerName, AmountDue, tax information, addresses, and invoice line items.

Because the model is pretrained to recognize common invoice structures, it can process invoices that use different layouts and formats while returning normalized structured output. This directly satisfies the requirement to extract invoice number, total, and vendor information from scanned documents while minimizing development and training effort.

The Azure AI Vision Read API performs OCR and can extract printed or handwritten text, but it does not provide invoice-specific semantic field extraction. Custom named entity recognition would require training and is designed for text entity extraction rather than document-layout understanding. A Document Intelligence custom neural model can handle organization-specific forms and layouts, but it requires custom model training, directly violating the requirement.

Therefore, the prebuilt invoice model provides the required combination of OCR, document understanding, structured extraction, and zero custom training.

Study Guide references: Azure AI Document Intelligence prebuilt invoice model; structured field extraction; OCR; invoice fields and line items.


Question 4

You need to secure an Azure OpenAI resource so that it is only reachable from your virtual network and not from the public internet.

What should you configure?



Answer : B

Configure an Azure Private Endpoint through Azure Private Link for the Azure OpenAI resource and disable public network access. A private endpoint assigns the Azure OpenAI service a private IP address from the selected virtual-network subnet, enabling workloads inside the VNet---or connected networks with suitable routing and DNS---to reach the service over private Azure networking rather than through its public endpoint. Microsoft explicitly documents private endpoints as the mechanism for accessing an Azure OpenAI resource after public network access is disabled.

For complete network isolation, creating the private endpoint alone is not sufficient if the resource continues accepting public connections. Microsoft security guidance recommends validating private endpoint connectivity and then disabling public network access, ensuring that permitted application traffic flows through Private Link. Private DNS should also be configured so the normal Azure OpenAI hostname resolves to the private endpoint address from within the VNet.

Azure AI Content Safety controls harmful-content detection rather than network accessibility. API key rotation improves credential hygiene but does not prevent public network connectivity. CORS controls browser-origin behavior and is not a network-isolation boundary.

Therefore, the required network architecture is Private Endpoint + Azure Private Link, with public network access disabled.

Study Guide references: Azure OpenAI networking; Azure Private Link; private endpoints; private DNS; public network access controls.


Question 5

You must ensure an Azure OpenAI-powered application never exceeds a defined token-per-minute budget across all users to avoid runaway costs.

What should you configure?



Answer : B

Configure deployment-level Tokens-per-Minute (TPM) and Requests-per-Minute (RPM) rate limits for the Azure OpenAI model deployment. Azure OpenAI quota is allocated to individual model deployments in TPM units, and Microsoft documents that the TPM assigned to a deployment directly maps to the rate limit enforced for inference requests against that deployment. An associated RPM limit is also enforced according to the model's TPM-to-RPM ratio.

Because all application users invoke the same deployment, this enforcement occurs at the deployment boundary rather than independently for each end user. Once request traffic would exceed the configured capacity, Azure applies throttling rather than allowing unrestricted token consumption. Microsoft Foundry additionally supports explicit token-management controls through AI Gateway, where a TPM limit can be set for a model deployment; requests exceeding the limit receive HTTP 429 Too Many Requests responses.

Content filtering governs safety categories, not consumption. An Azure AI Search vector-index limit controls search storage rather than generated tokens. Temperature changes sampling randomness and has no enforcement role in rate or cost control.

Study Guide references: Azure OpenAI quota management; TPM/RPM rate limits; deployment capacity; throttling; Foundry AI Gateway token management.

===============


Question 6

You are implementing an application by using Azure Event Grid to push near-real-time information to customers.

You have the following requirements:

* You must send events to thousands of customers that include hundreds of various event types.

* The events must be filtered by event type before processing.

* Authentication and authorization must be handled by using Microsoft Entra ID.

* The events must be published to a single endpoint

You need to implement Azure Event Grid

Solution: Publish events to a partner topic. Create an event subscription for each customer.

Does the solution meet the goal?



Answer : B

Detailed The proposed partner-topic design does not match the stated need for one publisher endpoint that can fan events into thousands of logically distinct customer topics. Event Grid domains are specifically designed to expose a single publishing endpoint while containing large numbers of topics, with subscriptions and filtering applied at the topic level. Partner topics are intended for partner-published SaaS events and do not substitute for the domain architecture required by this high-scale custom publishing scenario.

Study Guide Alignment: Azure service integration: Service Bus, Event Grid, Azure Functions triggers/bindings, and event-driven processing.

Official Microsoft Learn Reference: AI-200 Study Guide | Azure Event Grid event domains | Authenticate Event Grid publishing with Microsoft Entra ID


Question 7

You have an Azure Service Bus namespace with a partitioned queue named queue1.

You plan to send a large number of messages through queue1 over the next few weeks. The order of messages will be random. You must minimize the possibility of message transmission interruption by transient failures of individual partitions.

You need to use the optimal configuration of the partition key in the messages.

Which configuration should you use?



Answer : D


Page:    1 / 14   
Total 142 questions