Building Agentic AI Applications One d20 at a Time, with Java, Spring AI, and Model Context Protocol

Learn agentic AI architecture and design patterns through an unexpectedly illuminating lens: developing an AI Dungeon Master for playing D&D that actually works. I’ve played D&D with all the AIs for two years, and, now developed several Model Context Protocol (MCP) tools, prompts, resources, and other “servers” to make playing even better. D&D provides the perfect analog for enterprise systems - both involve complex rule sets, unpredictable actors, and the constant threat of critical failures at exactly the wrong moment. Learning how to add AI code to your apps is way more interesting when you’re fighting trolls instead of babysitting compliance workflows.

Using Java and Spring AI, this talk will cover: (1) MCP fundamentals and their role in agentic AI applications (2) How and why to use each MCP component (3) Spring AI patterns for rapid tool creation and modification (4) Handy development workflows and tools for agentic AI applications.

You’ll get a handle on how to write agents, and also how to start thinking differently when you weave AI into your apps.