Foundry Agent Service + Microsoft Agent Framework Explained

Move your AI agents from prototype to production using Microsoft Foundry Agent Service. Deploy directly from your local environment, run with secure identity and scoped permissions, and monitor every interaction so you can debug, improve, and scale without losing control. Publish agents into the tools your team already uses and ensure every action is traceable, governed, and isolated. Ground your agents in real work and business data to generate outputs that are actually useful. Pull from emails, meetings, and operational systems to create personalized insights, documents, and presentations. Build faster with familiar tools and frameworks, then manage performance, cost, and quality across all your agents as they scale. Jeff Hollan, Partner Director, AI Agent Services, shares how to operationalize AI agents across your organization—from deployment to real-world impact.  ► QUICK LINKS:  00:00 - Build single and multi-agentic workloads 00:44 - Build agents at scale with Foundry 01:33 - Demo: Sales meeting preparation agent 03:32 - How it works 04:48 - Access controls 05:44 - Publish the agent 06:23 - Direct integration with Microsoft 365 07:26 - Work IQ, Foundry IQ, & Fabric IQ 10:24 - Agent creation 11:21 - See what's happening in the code 12:54 - Manage performance<