Build and scale the next generation of AI agents in Microsoft Foundry | BRK240
Tina Schuchman, Hiroaki Kimura, and Jeff Hollan present an end-to-end agent workflow, showing how development, governance, and operational measurement fit together in Microsoft Foundry.
From local development to hosted agents
The session follows the journey from the first line of agent code to measuring business impact. It covers local development with GitHub and VS Code, followed by deployment to hosted agents in Microsoft Foundry.
Two parts of that workflow address how agents access tools and enterprise information:
- Toolboxes connect agents to governed tools.
- Foundry IQ grounds agents across enterprise data.
Measuring and improving agent performance
The session covers continuous improvement through tracing, evaluation, and optimization. It also examines cost and return-on-investment insights, bringing agent quality and economics into the same development lifecycle.
These activities lead into publishing agents to Microsoft 365 Copilot and Teams, where users can access them.
Session resources
This is an intermediate-level breakout session from Microsoft Ignite 2026, identified as BRK240 | Agent Platform.