Build context-aware agents: From data to decisions | BRK240
Amanda Silver and Marco Casalaina explain how to build context-aware AI agents by combining enterprise knowledge, business data, and work signals, using Foundry IQ, Fabric IQ, and Work IQ to add orchestration and governance so agents can act within trusted boundaries.
Overview
High-performance agents depend on an enterprise intelligence layer that can:
- Bring together context, enterprise data, orchestration, and governance
- Let agents search across organizational knowledge
- Enable agents to reason over business data
- Add awareness of people and work signals
- Keep actions within trusted boundaries so systems can scale reliably
The session focuses on how Foundry IQ, Fabric IQ, and Work IQ fit together to provide that intelligence layer for building scalable, reliable agent ecosystems.
Key themes covered
Why agentic projects struggle without context
- Agentic AI projects can fail or underperform when agents lack access to the right organizational context.
Foundry IQ: connecting agents to enterprise knowledge
- Foundry IQ is presented as a way to connect agents with enterprise knowledge so they can retrieve and use organizational context during decision-making.
Demo scenario: connecting an agent to enterprise intelligence
- The session transitions into a demonstration that connects a “refund agent” to enterprise context using Web IQ and Foundry IQ.
Fabric IQ: ontology generation and semantic integration
- Fabric IQ is shown generating an ontology and enabling semantic integration over business data.
Using an ontology as an MCP server
- The ontology is used as an MCP server and connected with Foundry IQ, positioning MCP as a mechanism for exposing structured context to agents.
Context delegation in data agents
- The presenters discuss context delegation as a pattern for data agents—how responsibility for retrieving/structuring context can be delegated so agents can reason and act more reliably.
Work IQ: connecting to organizational knowledge and work signals
- Work IQ is introduced as a way to connect agents to organizational knowledge and signals about people and work, so agents can operate with more situational awareness.
Secure, governed, scalable agent ecosystems
- The session closes by emphasizing building agent ecosystems that are:
- Secure
- Governed
- Scalable
- Reliable