Keynote - Its Up To Java Developers to Fix Enterprise AI

on YouTube ↗50m

" summary

Argues that enterprise AI succeeds only when engineering replaces ad hoc prompts and shows a Java-first solution: a server framework called imbabel that uses deterministic planning (goal oriented action planning driven by the Java type system), runs actions as Java methods so the planner can replan after each step, and mixes small LLM calls with regular code to reduce hallucination and cost. Demonstrates unfolding tools to avoid tool bloat, agentic RAG across Neo Lucene PGVector and file stores, and grounded agent memory linked to entities.

AI-written summary · about

" chapters (6)

1. 0:00Java developers in enterprise AI
2. 4:45Why enterprise GenAI fails
3. 13:59Fixing enterprise AI challenges
4. 21:50Java developers' role in GenAI
5. 28:15Inbabel's approach to GenAI
6. 49:44Key takeaways and resources

" related talks

• 44mJava Meets AI: Empowering Spring Developers to Build Intelligent AppsTimo Salm, Sandra Ahlgrimm · Jfokus 2026
• 27mEngineering Scalable AI with JavaNaci Dai · Java Day Istanbul 2026
• 53mUnleash the Power of AI Agents with Java Building Intelligent ApplicationJuarez Barbosa Junior · Build Stuff 2025
• 55m"You can't use AI directly in Java!" Hold my prompt...Przemyslaw Jakubowski · Java ON 2024
• 02h 23mJava for an AI World - KeynoteZhitao Li, Georges Saab, Mikael Vidstedt, Ana-Maria Mihalceanu, Lize Raes, Chad Arimura, Nicolai Parlog, Billy Korando, Ikroop Dhillon · JavaOne 2026
• 45mBuilding Intelligent Enterprises: AI-Powered Java with LangChain4j & CDIEmily Jiang · JCON EUROPE 2026
• 53mUnlocking AI's Potential: Smarter Coding in Apache NetBeans IDEGaurav Gupta · Great International Developer Summit 2025
• 10mAI agents on the JVMFrederik Pietzko · JCON EUROPE 2026