Driving agentic innovation w/ MCP as the backbone of tool-aware AI

In this technical deep-dive, Maria unveils how Microsoft is shaping the future of agent-tool interactions through the Model Context Protocol (MCP). As AI agents evolve from simple task runners to autonomous collaborators, MCP emerges as the preferred communication mechanism—enabling secure, scalable, and OS-agnostic tool orchestration across platforms like Microsoft Foundry, GitHub, and VS Code. Join the conversation on the Microsoft Foundry Discord: https://aka.ms/AIDevDays/Discord Sign up for an Azure Free Trial: https://aka.ms/AzureFreeTrialYT Get hands-on with today's MCP lab repo: https://aka.ms/AIDevDays/LabDay1 Catch up on-demand: https://aka.ms/AIDevDays/OnDemand Chapters 00:00 – Welcome & Introduction: Building Tool-Aware Agents with MCP 00:24 – What is an Agent? Deterministic vs. Non-Deterministic Programs 01:23 – Why Tools Matter: Retrieval, Actions, and Memory in AI Agents 02:09 – Enter MCP: Standardizing Tool Access Across Ecosystems 03:21 – MCP Explained: Secure, Unified Access for Agents and Assistants 04:02 – MCP Clients and Servers: How They Work Together 05:00 – Types of MCP Servers: Local vs. Remote and Their Use Cases 06:10 – Demo: Using MCP Servers for Automated Reporting and Testing 08:16 – Microsoft’s Bet on MCP: First-Party Servers and Integration Options 09:56 – Foundry Tools: Access 1,000+ Tools for Building Agents 10:44 – Managing MCP Servers: Registries and Enterprise Governance 13:16 – Demo: Deploying MCP Servers with Azure API Center 15:22 – Wrap-Up: From Code to Registry and Building Enterprise-Ready Agents