Practice Free AI-200 Exam Online Questions
A healthcare application uses Azure AI Language Text Analytics for Health. The team needs to extract medications and their associated dosages as structured output.
Which entity categories are returned?
- A . BodyStructure
- B . Frequency
- C . MedicationName
- D . Dosage
- E . MedicationRoute
An organization needs to implement Azure AI Language for analyzing customer support tickets. The solution must categorize tickets into 15 categories AND extract the product name and issue type from each ticket.
Which combination of Azure AI Language features is required?
- A . Custom single-label text classification for 15 categories
- B . Prebuilt NER for product and issue extraction
- C . Custom multi-label text classification for 15 categories
- D . Custom NER for product name and issue type extraction
- E . Key phrase extraction for product and issue identification
You are building a solution that uses Azure OpenAI with LangChain. You need to implement a ReAct agent that can search the web and query a SQL database.
What are the required components?
- A . LangGraph StateGraph for agent orchestration
- B . Tool definitions for web search and SQL query
- C . SerpAPI or Bing Search tool for web search
- D . Azure OpenAI ChatOpenAI as the LLM backend
- E . LangChain AgentExecutor with ReAct agent type
An AI application requires a globally distributed database with single-digit millisecond latency for storing user session vectors. You choose Azure Cosmos DB for NoSQL.
Which feature should you enable to ensure the application automatically reads from the nearest region?
- A . Service-managed failover
- B . Multi-region writes
- C . Availability Zones
- D . Integrated Cache
An AI workload on AKS requires a large amount of temporary disk space (500GB) for processing video files. You want the fastest possible disk I/O and do not need the data to persist if the node is deleted.
Which disk type should you use?
- A . Azure Managed Disks (Standard HDD)
- B . Ephemeral OS disk or Local SSD
- C . Azure Archive Storage
- D . Azure Managed Disks (Premium SSD)
An AI application uses Azure Functions to process high volumes of telemetry data. You notice the function is being throttled. You want the function to scale based on the number of messages waiting in an Azure Queue.
Which scaling component is used by the platform?
- A . The Load Balancer
- B . The Traffic Manager
- C . The Scale Controller
- D . The App Service Plan
An AI agent architecture requires a ‘Dead-Letter’ queue for messages that cannot be processed after 5 attempts.
Which service provides this built-in capability?
- A . Azure Load Balancer
- B . Azure Event Grid
- C . Azure Front Door
- D . Azure Service Bus
You are deploying a global AI application. You want to use an Azure Cosmos DB configuration that allows for the lowest possible write latency by having multiple writable replicas around the world.
What should you configure?
- A . LRS (Locally Redundant Storage)
- B . Strong Consistency
- C . Multi-region writes
- D . Serverless mode
A developer is integrating Azure OpenAI Service into an enterprise chatbot. The solution must ensure that responses do not contain harmful content and must log all interactions for compliance.
Which two components should be configured?
- A . Use Azure Monitor to log API calls to a Log Analytics workspace
- B . Configure Responsible AI dashboard in Azure Machine Learning
- C . Deploy the model to a consumption-based endpoint
- D . Enable Azure OpenAI content filtering policies
- E . Enable system prompt injection via the API
An Azure AI Vision custom model is trained using Azure AI Studio’s Vision Studio. The model must detect defects in manufacturing images with 95% precision. The trained model achieves 88% precision on the test set.
What actions improve precision?
- A . Add more training images of defect types with low precision
- B . Review and relabel ambiguous training images
- C . Use a pre-trained Azure AI Vision model instead
- D . Adjust the probability threshold upward to increase precision at the cost of recall
- E . Reduce the number of defect classes