Beyond local tools: Deep dive into Agentic AI patterns with Spring AI

Building production-grade agentic systems isn’t just about choosing the right protocols: It’s about designing how agents collaborate, make decisions, and stay aligned with your goals.
When should agents communicate peer-to-peer versus through a supervisor? How do you balance autonomy with controllability?

In this talk, we’ll go beyond local tools to focus on the orchestration layer: agent communication patterns, supervision strategies, memory and state management, and the architectural trade-offs that determine whether your multi-agent system is robust or fragile.
Using Spring AI, we’ll walk through concrete design patterns and live demos from simple single-agent setups to coordinated multi-agent workflows.