As AI becomes more embedded in the tools we use daily, organizations need ways to securely and efficiently expose internal systems to intelligent agents. This session introduces the Model Context Protocol (MCP) — an open framework for surfacing functional, code-level tools to AI systems. We’ll walk through how MCP servers work, how they expose discrete capabilities as callable tools, and how these tools can be discovered and executed by AI agents like Microsoft Copilot as part of complex prompt execution.

Through real-world examples and live demos, we’ll show how to build and deploy an MCP server using C#, structure it to expose services relevant to your business, and register it for use by AI agents in environments such as Microsoft Teams, Visual Studio Code, and Azure DevOps. Attendees will leave with a clear understanding of the architecture, security model, and practical use cases for MCP — along with the confidence to begin connecting their own tools and services into the growing AI ecosystem.

Topics and technologies covered include:
• Implementing and deploying an MCP server in C#
• Creating Agents in Copilot Studio and Azure AI Studio Foundry
• Integrating a hosted MCP server into those agents
• Deploying agents in Microsoft Teams for easy access to end users