TechWorkshop L300: MCP and AI Gateway

This lab teaches you how to design and build AI applications using the Model Context Protocol (MCP) and Azure AI Gateway. You will learn how to create MCP servers that connect AI models to external tools and data sources, configure Azure API Management as an AI Gateway to manage and secure AI API traffic, and build end-to-end solutions that leverage MCP for tool integration and AI Gateway for centralized AI endpoint management, monitoring, and governance.

Exercises

This lab has exercises on:

• Deploy and configure resources
• Build an MCP server with tool integrations
• Configure Azure API Management as an AI Gateway
• Connect MCP clients to AI Gateway endpoints
• Monitor and manage AI traffic through the gateway
• Secure and govern AI endpoints
• Resource cleanup

The lab is available as GitHub pages here.

Prerequisites

For running this lab you will need:

• An MCAPS subscription with access to Microsoft Foundry. External MCAPS subscriptions are preferred, though internal MCAPS subscriptions should work as well. Microsoft EMU accounts will not have sufficient privileges to perform all of the actions necessary.
• A desktop, laptop, or virtual machine and access to install software on that machine.