Discover how to take Model Context Protocol (MCP) servers out of local development and scale them securely in the cloud. We will demonstrate how to deploy and manage MCP tools using Cloud Run and GKE while locking them down with Identity-Aware Proxies and IAM best practices. Finally, you'll learn how to connect your secure MCP infrastructure directly to the Agent Platform to power high-traffic production workloads.

What we’ll cover:

• Cloud Run: Serverless, secure-by-default MCP deployment.

• GKE: Scaling your MCP tools for complex, high-traffic workloads.

• Agent Platform: Wiring your secure MCP servers directly to your AI agents.

• Security: IAM, identity-aware proxies, and best practices for locking down your tools.