Build Generative AI Solutions with Microsoft Foundry AI-102 Practice Questions (Page 8)

Questions 36–40

1. 36foundation · easy

   What is the primary purpose of a prompt flow in Microsoft Foundry?

   • ATo train custom models
   • BTo store prompt templates
   • CTo orchestrate a sequence of steps that interact with models and tools
   • DTo monitor model performance
   Reveal Solution Discussion Select an answer first
2. 37foundation · easy

   Which Azure service is commonly used in a RAG pattern to index and search documents?

   • AAzure Functions
   • BAzure Blob Storage
   • CAzure Logic Apps
   • DAzure AI Search
   Reveal Solution Discussion Select an answer first
4. 38application · medium

   A logistics company has a deployed prompt flow in Microsoft Foundry that processes shipment tracking queries. They want to integrate this flow into their mobile app. The app needs to send a query and receive a response. They also want to monitor the flow's performance. Which approach should they take?

   • AUse the Foundry SDK in the mobile app to call the deployed flow endpoint, and use Foundry's monitoring tools to track performance.
   • BUse a REST API call to the flow endpoint, but skip monitoring to reduce complexity.
   • CUse the Azure portal to manually invoke the flow for each app request.
   • DRe-implement the flow logic in the mobile app to avoid the SDK dependency.
   Reveal Solution Discussion Select an answer first
5. 39foundation · easy

   Which model type is BEST suited for a use case that requires generating human-like text responses in a chat application?

   • ASpeech recognition model
   • BComputer vision model
   • CAnomaly detection model
   • DLarge language model (LLM)
   Reveal Solution Discussion Select an answer first
6. 40expert · hard

   A government agency is planning a generative AI solution for internal document summarization. They have strict data residency requirements that mandate all data processing occur within a specific Azure region. They also have a limited budget and need to minimize costs. They are considering using Microsoft Foundry. Which planning decision should they make?

   • ADeploy the hub in the required region but use a model from a different region to get better performance.
   • BDeploy the Foundry hub and all resources in the required region, and use a model that is available in that region.
   • CDeploy the hub in a different region to save costs, and use a global model endpoint.
   • DUse a serverless model API that may process data in multiple regions.
   Reveal Solution Discussion Select an answer first