MCP has quickly become the standard for connecting AI agents to real-world tools and services. But until recently, MCP was limited to local development. This talk covers the shift from local to remote MCP servers, including the technical challenges of deploying them at scale and recent transport layer improvements that address these challenges.I'll share best practices for building reliable MCP servers based on learnings from working with customers and internal teams, and explore the benefits of building a stateful MCP server. Finally, the session will include a live demo showing how to deploy a remote MCP server to production.