[vJUG CONNECT] Agentic Memory & Structured AI: Scaling Smarter Java Development

Register: https://www.meetup.com/virtualjug/events/310960485/

We’ll kick things off with Simon Maple exploring Spec Driven Development, showing why structured prompts are the key to scaling AI coding. Then, Andreas Kollegger, Lead for GenAI Innovation at Neo4j, will dive into Agentic Memory – Ripples of the Matrix, a thought-provoking exploration of how memory in time and space shapes the way agents collaborate, compete, and evolve. Finally, Holly Cummins will demonstrate Using LangChain4j to Make Your Java Apps Smarter, showing how to harness AI frameworks to integrate LLMs with real-world Java applications.
Whether you’re tuning in from around the world on YouTube or joining us in person at Tessl HQ with the LJC, expect insightful talks, lively discussion, and of course, pizza and beer for those in London!

Session: Agentic Memory - Ripples of the Matrix by Andreas Kollegger

The first drop seem innocent, refreshing almost. An agent's chain-of-thought descends from the sky, impacting the surface of our reality. Then more drops. Puddles form and ripples become interference patterns. Eventually, we're swimming in a digital sea. But seriously, let's talk about two dimensions of memory: time & space - working, short-term, long-term memory - single-agent, multi-agent, distributed scopes - collaborative, misaligned, competitive goals