St. Louis Cloud and AI Summit Monthly

This will be an in-person-only event.

Main Talk:
The Model Context Protocol (MCP) has quickly become the standard for connecting large language models to the systems developers actually work with. This session is a practical, code-forward introduction for developers who want to understand what MCP is and how to build servers that integrate cleanly across the modern AI toolchain. Using working examples in C# and TypeScript, we will cover the architecture, design tradeoffs, and security patterns that separate a solid demo from a production-ready server.

We will also ground the material in reality with examples of how ArchitectNow and our clients use MCP servers today, plus a rundown of commercial servers you can adopt immediately.
Key topics:

Core MCP concepts and how servers integrate with Copilot, GitHub Copilot, Claude Code, Claude Desktop, and custom apps via the Microsoft Agent Framework
Local (stdio) vs. hosted/remote (streaming HTTP) architectures and when to use each
Designing and describing MCP "Tools," with sample implementations in C# and TypeScript
APIs vs. MCP: how they differ, and how to expose existing APIs through MCP wrappers
Delegated security and safe authorization patterns
Real-world ArchitectNow examples and a list of ready-to-use commercial MCP servers

Attendees will leave with a clear understanding of MCP architecture, the design and security tradeoffs that matter in production, and a practical mental model for how these servers fit into real AI toolchains.

Agenda...
5:30pm Central Time - Networking
6:00pm - Announcements
6:10pm - Main Talk
7:15pm - Closing