Plan and manage an Azure AI solution 1.1 Choose the appropriate Foundry services for generative AI and agents
A travel agent built in Foundry Agent Service must (1) remember each traveler's seat and…
Study guide skill: Choose appropriate memory, tool, and knowledge integration services for agent solutions
Question B2-001
A travel agent built in Foundry Agent Service must (1) remember each traveler's seat and meal preferences across future sessions, (2) ground answers on the company's curated travel policy, and (3) search an itinerary PDF that the traveler uploads during the current chat. Which combination should you use? (Memory in Foundry Agent Service is currently in preview.)
Select one answer.
Answer: C
Microsoft's guidance is to use memory (preview) for user-specific context that is learned and persists across interactions, a Foundry IQ knowledge base to ground the agent on curated organizational content, and the file search tool for documents the user provides during an interaction. Memory extracts and consolidates knowledge from conversations, so it isn't a substitute for a curated policy source or for searching a file uploaded in the current chat.
Deep dive
How it works
The memory page gives a three-way rule for this exact design. Use memory to capture user-specific or agent context that is learned and persists through interactions over time. Use a Foundry IQ knowledge base to ground your agent on curated organizational content. Use the file search tool to search user-provided documents during an interaction.
Map the travel agent's needs to that rule:
- Seat and meal preferences across future sessions: memory. Memory in Foundry Agent Service is designed for long-term memory. It extracts meaningful information from conversations, consolidates it into durable knowledge, and makes it available across sessions. User profile memory holds durable preferences such as language preference or accessibility needs.
- Curated travel policy: a Foundry IQ knowledge base. A knowledge base connects to your data stores, and Foundry IQ uses agentic retrieval to retrieve relevant information, enforce user permissions and return grounded answers with citations.
- Itinerary PDF uploaded in this chat: the file search tool, which retrieves relevant information from your documents, including user-provided documents.
Memory is in preview, so treat its behavior as subject to change.
The other options, and when you'd use them
- File search for preferences, memory for policy, knowledge base for the upload: Every mapping is reversed. Preferences are learned user context (memory), the policy is curated organizational content (knowledge base), and the upload is a user-provided document (file search).
- Knowledge base for preferences, memory for policy, file search for the upload: A knowledge base is for curated organizational content, not for preferences learned from each traveler. And memory extracts information from conversations, so it isn't the source of truth for a curated policy.
- Memory for all three: Memory extracts key information from conversations, such as preferences and facts, and consolidates it. It doesn't replace curated grounding or a search over a document the user just uploaded.
Exam tip
Learned about the user, across sessions = memory. Curated by the organization = Foundry IQ knowledge base. Uploaded right now = file search.
References
Last reviewed: