AI-200 Practice Test 27 — 15 Questions

Question 1

You are designing an AI application architecture where embeddings are generated on-the-fly by an Azure Function and stored in Azure Cosmos DB for NoSQL. During peak load, the Azure Function times out due to rate limits (HTTP 429) from Azure OpenAI Service. What design pattern should you implement to make the ingestion pipeline resilient?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Implement an asynchronous message queue using Azure Service Bus to buffer ingestion requests and handle rate-limit retries. Implementing an asynchronous messaging pattern using Azure Service Bus or Event Hubs decouples the client ingestion request from embedding generation, allowing rate limiting and retries.

Answer

Increase the Azure Function execution timeout limit to 24 hours

Answer

Disable SSL verification on the Azure OpenAI client connection

Answer

Switch from Azure OpenAI to local text file storage

Question 2

You are troubleshooting a Python application that fails to connect to Azure Key Vault using a managed identity. You want to inspect the exact exception stack trace and identify whether the managed identity token request failed. Where should you look in your local development or cloud diagnostics?

Accepted Answer

Application Insights exceptions and traces tables in the Log Analytics workspace. Application Insights logs exceptions and dependency tracking, capturing authentication failures when attempting to acquire tokens for Azure services.

Answer

Azure Security Center threat detection alerts

Answer

Azure Key Vault access policies audit logs

Answer

Azure Advisor cost optimization recommendations

Question 3

You are developing a RAG solution and need to evaluate the quality of your vector retrieval results from Azure AI Search. Which metric is commonly used to measure the proportion of relevant documents retrieved in the top-k results?

Accepted Answer

Recall@k. Recall@k measures the proportion of relevant items found in the top-k retrieved results relative to all relevant items.

Answer

HNSW efSearch latency

Answer

Request Units per second (RU/s)

Answer

BM25 term frequency

Question 4

You are configuring semantic ranking in Azure AI Search. You need to specify which fields in your index contain the title and content that the semantic ranker should analyze. Where do you configure these field mappings?

Accepted Answer

In the semantic configuration section of the index schema. Semantic configuration in Azure AI Search defines the prioritized fields for title, content, and keywords used by the semantic ranker.

Answer

In the indexer skill set cognitive map

Answer

In the Azure OpenAI model deployment properties

Answer

In the vector profile definition

Question 5

An enterprise AI application experiences intermittent HTTP 429 (Too Many Requests) errors when calling Azure OpenAI Service during peak hours. You need to analyze the exact token consumption trends and throttling patterns over the past week. Where should you look first in Azure Monitor?

Accepted Answer

Azure Monitor Metrics explorer, charting 'Processed Tokens' and 'Token Generation' metrics filtered by deployment name. Metrics explorer in Azure Monitor allows you to query metrics such as 'Processed Prompts' and 'Tokens' for Azure OpenAI, categorized by dimensions like model deployment name, to diagnose throttling.

Answer

Log Analytics workspace querying the AzureActivity table for throttling events

Answer

Azure Cost Management and Billing dashboard for daily spend trends

Answer

Application Insights availability tests configured with a 5-minute frequency

Question 6

You are designing a high-throughput RAG pipeline where documents are continuously ingested and queried. You choose Azure Cosmos DB for NoSQL as your vector store. How does Cosmos DB handle vector indexing updates when new documents are inserted?

Accepted Answer

Vector indexes are updated asynchronously in the background by the database engine. Azure Cosmos DB indexes vector embeddings asynchronously in the background as documents are written, balancing write performance with index availability.

Answer

You must manually trigger an index rebuild stored procedure every hour

Answer

Vector search is completely unavailable until the entire collection is re-indexed from scratch

Answer

New documents cannot be queried using VectorDistance until the container is restarted

Question 7

Your AI application is hosted on Azure App Service and connects to Azure Cosmos DB and Azure OpenAI. You want to ensure that if credentials or connection strings ever leak, they cannot be used from unauthorized networks. Which feature should you implement?

Accepted Answer

IP firewall rules and virtual network service rules on the backend services restricting access to App Service. Network service tags and IP firewalls on Azure Cosmos DB and Azure OpenAI restrict access to only allow incoming traffic from the specific outbound IP addresses or virtual network subnet of the App Service.

Answer

Disabling managed identities across all application components

Answer

Configuring Azure Front Door with SSL offloading only

Answer

Enabling public anonymous access with client-side rate limiting

Question 8

You are designing a data architecture on Azure where embeddings are generated offline by an Azure Databricks pipeline and stored in Azure AI Search. You want to minimize ingestion latency and optimize throughput when uploading millions of pre-computed vectors. Which client library API pattern should you use?

Accepted Answer

Use SearchClient batch upload APIs with up to 1000 documents per batch. When uploading large batches of documents containing pre-computed vectors to Azure AI Search, using the SearchClient.UploadDocumentsBatch (or mergeOrUpload) method with batched payloads of up to 1000 documents maximizes throughput.

Answer

Stream vectors directly through the Azure AI Search integrated vectorizer skillset

Answer

Send documents one-by-one using synchronous CreateOrUpdate API calls

Answer

Execute direct SQL INSERT statements against the underlying search partition nodes

Question 9

You are designing an IoT data ingestion architecture using Azure Event Hubs. Devices occasionally send malformed JSON payloads that cause stream processing jobs to crash. You need to route these invalid messages to a separate error container without stopping the main ingestion stream. What should you implement?

Accepted Answer

Use Azure Stream Analytics with an error output data stream configured for malformed events.. Event Hubs does not have a native bad-record-routing queue per se, but stream processing engines like Azure Stream Analytics or custom consumers can implement try-catch error handling patterns and output invalid records to an error sink.

Answer

Enable Event Hubs Capture with the JSON validation flag set to strict.

Answer

Set up an Event Grid system topic filter for 'MalformedEvent'.

Answer

Configure the Service Bus dead-letter queue on the Event Hub producer client.

Question 10

Your data science team uses Azure Machine Learning notebooks. To prevent data exfiltration, you need to ensure that users cannot download datasets or trained model artifacts directly from the workspace to their local corporate laptops. Which feature should you configure?

Accepted Answer

Workspace configured with a managed virtual network and outbound rules blocking public internet egress, plus disabled local file downloads in compute instance settings. Disabling public internet access and configuring data egress restrictions or using compute instances with outbound internet access disabled prevents data downloading to local devices.

Answer

Using Azure Front Door with Web Application Firewall disabled

Answer

Disabling Azure Monitor diagnostic logging

Answer

Assigning the Contributor role to all external collaborators

Question 11

You are optimizing an enterprise Azure Cosmos DB for NoSQL vector database. Which THREE administrative or architectural practices help ensure high availability and predictable performance? (Choose three)

Accepted Answer

Choose a partition key that evenly distributes data and query load across physical partitions. Ensuring high performance and availability in Cosmos DB involves configuring consistent request unit provisioning, setting up multi-region writes, and defining effective partition keys.

Answer

Store all documents in a single partition without a partition key

Answer

Disable automatic indexing entirely across all properties

Question 12

When configuring an Azure AI Search index, which TWO field attributes are required if a field is going to be used for full-text keyword search and filtering? (Choose two)

Accepted Answer

searchable. To search and filter text fields, the field must be marked as searchable (for BM25 search) and filterable (for OData filters).

Answer

vectorSearchProfile

Answer

facetable

Answer

sortable

Question 13

You are designing a high-scale vector search architecture using Azure AI Search. Which THREE operational and architectural best practices should you follow to ensure optimal performance and cost-efficiency? (Choose three)

Accepted Answer

Balance replica and partition counts based on your query throughput and index size requirements. Best practices for Azure AI Search include monitoring resource utilization, right-sizing replicas and partitions, and leveraging vector compression (scalar quantization).

Answer

Store all raw PDF files inside the vector float32 array fields

Answer

Deploy all indexes on the Free tier to eliminate infrastructure costs

Question 14

When configuring vector search in Azure AI Search, which TWO parameters must be defined when creating a vector profile? (Choose two)

Accepted Answer

algorithm. A vector profile requires a profile name, an algorithm configuration name, and optionally a vectorizer name.

Answer

sqlConnectionString

Answer

vmSize

Answer

cosmosDatabaseName

Question 15

Your application makes frequent calls to Azure AI Document Intelligence. You need to ensure that if a transient network failure or HTTP 503 service unavailable error occurs, the Python SDK automatically retries the failed requests with exponential backoff. How should you initialize the client to support robust retry behavior?

Accepted Answer

Instantiate the client with custom retry policy parameters (e.g., retry_total, retry_backoff_factor). Azure SDK clients include built-in retry policies that can be configured with custom retry intervals, backoff factors, and maximum retry counts during client instantiation.

Answer

Write a custom infinite while loop wrapping every API call without sleep intervals

Answer

Disable all exception handling so the application crashes immediately on error

Answer

Configure Azure Traffic Manager with failover routing priorities