What if your AI assistant could search your Microsoft 365 files—just like a human coworker—with natural language? With Model Context Protocol (MCP), that vision is now real. In this session, we’ll explore how to build lightweight, plugin-friendly tools that bridge AI and enterprise data using Microsoft Graph. You’ll get a behind-the-scenes look at a real MCP tool that connects to OneDrive and powers intelligent file search in platforms like Claude and ChatGPT.

We’ll cover how to build, publish, and test your own MCP server, handle fallback behavior when Graph queries fail, and connect your plugin to AI assistants. Whether you're new to MCP or looking to take your integrations to the next level, this session is packed with practical takeaways and code you can use today.

You will learn:

• Understand the Power of MCP: Learn how Model Context Protocol works and why it's the key to connecting AI assistants with Microsoft Graph.
• Build Your Own AI-Powered File Search: Walk through how to write, publish, and test an MCP plugin that searches OneDrive files using Graph API.
• Deploy and Integrate Seamlessly: Discover how to handle query fallback, deploy your plugin to npm, and use it in real-world assistants like Claude.