AI-103: Developing AI Apps and Agents on Azure

By Microsoft · Question Mode

QuestionQ47

VerifiedImplement generative AI and agentic solutions

You have a chat application in a Microsoft Foundry project and a vectorized index in Azure AI Search.

You need to connect to the index so that the solution meets the following requirements:

  • Complex questions must retrieve information from multiple chunks.
  • Multi-turn conversations must influence retrieval planning.
  • Retrievals must run in parallel to reduce latency.

Which retrieval approach should you use?

Explanation

Agentic retrieval in Azure AI Search uses an LLM-backed knowledge agent that decomposes a complex user question into multiple focused subqueries, incorporates conversation history into query planning, executes the subqueries in parallel against the underlying knowledge source(s), and then semantically reranks and merges the results — satisfying multi-chunk retrieval, conversation-aware planning, and reduced latency at the same time. Classic RAG issues a single retrieval call per turn with no decomposition or parallelism, iterative retrieval performs sequential (not parallel) retrieval rounds, and chain of thought is a prompting technique unrelated to retrieval architecture.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!