Which Knowledge Type for Foundry Multi-Agent Docs?

AI-500 By islever Editorial Last updated

Design and implement agent memory, context management, and knowledge integration Specify technology components to include in a multi-agent solution

Short answer

C Use an Azure AI Search index as the single maintained knowledge layer, returning semantic answers with citations and query details for natural-language questions.

Question

You need to recommend a knowledge integration design for a Microsoft Foundry multi-agent solution. The agents answer questions by using shared documentation. The solution must meet the following requirements:
Updates must be available from a single maintained knowledge layer.
Retrieval responses must include citations and query details.
Content must support natural-language queries.
Users will ask the agents complex conversational questions. The questions will include follow-up context and terminology that does NOT always match the wording in the documentation.
Which knowledge type should you recommend?
  1. File
  2. Fabric IQ (OneLake Catalog)
  3. Azure AI Search Index Correct Answer
  4. Work IQ

Key Takeaways

The test is matching retrieval requirements (single maintained layer, citations plus query details, natural-language and vocabulary-mismatch queries) to a knowledge type; the trap is choosing a catalog or work-data layer because it sounds like a unified knowledge store.

This AI-500 question asks which Microsoft Foundry knowledge type best serves agents answering questions from shared documentation with citations, query details, natural-language matching, and a single maintained layer. An Azure AI Search index (C) satisfies all four requirements through hybrid/semantic retrieval, synonym and vector matching, and answer captions that carry query context and document citations.

Common mistake: The most common wrong pick is Work IQ (D), chosen because its name implies one maintained knowledge layer, but Work IQ surfaces Microsoft 365 work data (mail, chats, files) rather than providing retrievable shared documentation with query answers, highlights, and per-document citations.

Expert Analysis

Why the Answer Is Correct

An Azure AI Search index is a single, centrally maintained knowledge layer that ingests the shared documentation once and serves every agent from that one source, so content updates propagate everywhere without re-uploading files. Its semantic ranker plus vector/hybrid search returns answers with captions and highlights alongside each retrieved document's title and URL, which directly satisfies the requirement that responses include citations and query details. Because ranking is driven by embeddings and a semantic ranker rather than literal token overlap, conversational follow-ups and terminology that never appears verbatim in the documentation still retrieve the right passages. Synonym maps and language understanding further close the vocabulary gap between how users ask and how the docs are written.

Why the Other Options Are Wrong

A File knowledge type attaches documents directly to an agent: every update requires a new upload or re-index, there is no shared single maintained layer across agents, and raw file grounding yields no structured query answers or citation metadata. Fabric IQ (OneLake Catalog) is a semantic catalog for discovering and describing lakehouse tables, models, and lineage across a Fabric estate — it is built for data discovery and governed access, not for returning passage-level answers with captions over prose documentation. Work IQ is the Microsoft 365 context layer over an individual's work data (email, chats, files in M365); it does not host the shared, centrally maintained documentation set or emit the query-answer/citation payload this design requires.

Community Comment Notes

The single community contribution directly selects "C. Azure AI Search Index" and simply restates that option, offering no counter-argument for Fabric IQ or Work IQ. With no vote distribution and no dissenting comments available, the corpus gives no evidence contradicting the requirement-by-requirement match between Azure AI Search and the stated retrieval needs.

Official Reference

These links open the official vendor documentation in a new tab.

Exam Strategy

Read the four requirements as a checklist and score each option against all of them — a choice that satisfies only 'single maintained layer' but not citations/query details is a distractor. In Foundry knowledge questions, reserve 'IQ' options for Microsoft 365 or Fabric data-context scenarios and default to Azure AI Search when passage-level citations and natural-language retrieval are named.

Frequently Asked Questions

Why not use a File knowledge type for the shared documentation?

File grounding requires re-uploading content per agent, breaks the single maintained layer, and does not return structured query answers or citation metadata with each response.

How does Azure AI Search handle wording that doesn't match the documentation?

Vector/hybrid search plus the semantic ranker and synonym maps match meaning rather than exact tokens, so conversational follow-ups and unfamiliar terminology still retrieve relevant passages.

More AI-500 FAQ →

Discussion 1 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 1

  1. nihil94 👍 1 Answered C
    C. Azure AI Search Index

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Log in to comment, report an error, or add a note about this question.

Submitted for moderation before publishing. Keep it helpful and respectful.

Which Evaluator Isolates RAG Retrieval Failures per Agent

S-grade

Which Identity Access Configs for Foundry Agent Tools?

S-grade

Yes/No: Multi-Agent Observability Monitoring in Microsoft Foundry

S-grade

How Should You Configure Foundry Multi-Agent Prompt Evaluations?

A-grade

← Back to AI-500 Study Guide