AI-200 Exam Dumps : Developing AI Cloud Solutions on Azure

Verified Answer: 1) Instrument the application code with the OpenTelemetry SDK; 2) configure an Azure Monitor trace exporter in the OpenTelemetry SDK; 3) redeploy the instrumented services.

Detailed Explanation: Trace correlation requires the application to create OpenTelemetry spans and an exporter to send those spans to Azure Monitor. Instrumentation must therefore be present before the updated service is deployed. The previous Application Insights SDK instrumentation does not satisfy the case-study requirement that all tracing use OpenTelemetry. Configuring an OpenTelemetry view is unrelated to basic distributed-trace export, and calling TrackEvent is an Application Insights SDK pattern rather than the required OpenTelemetry approach.

Study Guide Alignment: Security and operations: Key Vault, App Configuration, managed identity, OpenTelemetry, Azure Monitor, and KQL-based troubleshooting.

Official Microsoft Learn References: AI-200 Study Guide | Enable Azure Monitor OpenTelemetry

Initialize the connection by using the account endpoint and key: CosmosClient

Define shared throughput: Database

Perform item CRUD operations: Container

The Azure Cosmos DB for NoSQL .NET SDK uses a clear resource hierarchy. CosmosClient is the top-level client object and is responsible for connecting to the Azure Cosmos DB account. Microsoft identifies it as the primary SDK class for accessing account-level resources and databases. It can be initialized by using the account endpoint together with an account key or another supported credential.

A Database represents a database resource within the account. Shared provisioned throughput is configured at the database level , allowing multiple containers within that database to share the same RU/s allocation. Microsoft documents that setting throughput on the database creates a shared-throughput database.

A Container is the SDK object used to interact with items stored in a Cosmos DB container. Microsoft specifically describes the Container class as the object used for item operations such as creating, reading, updating, deleting, and querying documents.

The Indexing policy option is not used for any of these requirements. It controls how Cosmos DB indexes document properties for query execution, not connectivity, shared throughput, or CRUD access.

Study Guide references: Azure Cosmos DB for NoSQL .NET SDK → CosmosClient; Database; shared throughput; Container; item CRUD operations.

Use Azure AI Search integrated vectorization with a skillset . Integrated vectorization automates the main ingestion stages required for RAG: an Azure AI Search indexer can read documents directly from a supported data source such as Azure Blob Storage, extract and split document content into chunks, invoke an embedding model, and store both searchable text and vector representations in the search index. Microsoft specifically states that integrated vectorization reduces development and maintenance effort because developers do not need to implement the chunking and embedding pipeline manually.

A skillset can include a Text Split skill for chunking and an Azure OpenAI Embedding skill or another supported embedding skill for vector generation. At query time, a matching vectorizer can convert user questions into vectors so Azure AI Search can perform similarity or hybrid retrieval for grounding the language model.

Custom text classification solves a different NLP problem. Building Azure Functions to process all 50,000 PDFs manually would introduce unnecessary custom orchestration and maintenance. Azure Table Storage does not provide the native vector-search and RAG indexing capabilities required here.

Study Guide references: Azure AI Search → integrated vectorization; indexers; skillsets; Text Split skill; embedding skills; vector indexes; RAG retrieval.

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