Practice Free AI-200 Exam Online Questions
You need to stream responses from Azure OpenAI to reduce perceived latency in a chat UI.
Which implementation details are required?
- A . Set stream: true in the API request body
- B . Buffer the entire response before rendering
- C . Use the delta field in each chunk to extract partial content
- D . Set a streaming-compatible Content-Type header in the response
- E . Process Server-Sent Events (SSE) on the client side
An AI workload on AKS is experiencing high costs due to idle GPU resources. You want to use a specific type of node that is significantly cheaper but can be reclaimed by Azure at any time.
Which node type should you add to your cluster?
- A . System node pool
- B . Premium node pool
- C . User node pool
- D . Spot node pool
You are building a serverless AI API using Azure Functions. You need to ensure the function can scale out to 100+ instances instantly to handle a burst of AI inference requests.
Which hosting plan is recommended for this ‘Elastic’ scaling?
- A . Free Plan
- B . Consumption Plan
- C . Basic Plan
- D . Dedicated Plan
An AI agent needs to ‘Call a Tool’ to check a customer’s credit score. The tool is a secure REST API.
How does the agent receive the schema of the tool to know what parameters to send?
- A . In a ‘tools’ definition block in the API call
- B . By guessing the JSON keys
- C . Through the system prompt
- D . In a separate PDF file
You are deploying a highly available chatbot backend using Azure Container Apps that calls an Azure OpenAI Service model. During high-traffic events, the Azure OpenAI endpoint responds with HTTP 429 (Too Many Requests). You want to configure the Container App to scale up its replica count automatically when outbound HTTP errors or high concurrent request volumes occur, using KEDA metrics.
Which two scaling configurations or rules should you implement? (Select TWO)
- A . Deploy an Azure Service Bus queue to intercept all synchronous HTTP payload streams.
- B . Define a scale rule based on CPU and memory utilization thresholds within the Azure Container Apps configuration template.
- C . Configure a custom KEDA scale rule using the ‘http’ scaler or an ‘azure-monitor’ scaler tracking request metrics.
- D . Configure an Azure AI Search synonym map to handle backend container thread pools.
- E . Set up an Azure API Management gateway in front of Azure OpenAI and use its metrics to trigger the KEDA scale rule.
You are designing a RAG pipeline where the source documents are 1,000-page PDFs. You want to ensure the ‘Chunks’ sent to the LLM contain enough context to be understood.
Which chunking strategy is most effective?
- A . Whole document chunks
- B . Single page chunks
- C . Fixed-size (e.g. 500 characters)
- D . Overlapping recursive character splitting
You are building a ‘Responsible AI’ dashboard. You want to track the number of times the model’s output was ‘Blocked’ by a content filter.
Which service provides this telemetry?
- A . Application Insights (via Azure OpenAI Diagnostic Logs)
- B . Azure Advisor
- C . Azure Policy
- D . Cost Management
An AI agent processes video files stored in Azure Blob Storage. You want the agent to start processing immediately after a new file is uploaded.
Which trigger is the most efficient and low-latency?
- A . Polling the storage API
- B . A ‘For Each’ loop in a Logic App
- C . An Event Grid trigger (Blob Created)
- D . A scheduled timer every 1 minute
You are deploying an AI model to Azure Container Apps. You want to restrict the container’s permissions so it can *only* write to a specific ‘Logs’ container in Azure Blob Storage.
Which RBAC role should you assign to the Managed Identity?
- A . Storage Account Contributor
- B . Storage Blob Data Contributor
- C . Storage Blob Data Reader
- D . Owner
You are developing an intelligent chat assistant that uses Azure Cosmos DB for NoSQL to store vector embeddings of product items. To support highly accurate vector similarity searches across millions of documents, you must configure the container’s indexing policy to include vector indexes.
Which two operations or configurations must you perform? (Select TWO)
- A . Configure an Azure Database for PostgreSQL extension inside the Cosmos DB account settings.
- B . Define a vectorEmbeddingPolicy on the container layout specifying the path, data type, distance function, and dimensions of the vector.
- C . Deploy an Azure Service Bus topic to dynamically flatten the vector float array into raw string tokens.
- D . Include a spatial index path pointing to the coordinates field using GeographyPoint mapping.
- E . Update the container’s indexingPolicy to define a vector index type (such as ‘diskANN’, ‘quantizedFlat’, or ‘flat’) on the specified vector path.