Foojay Podcast #86: Agents, MCP, and Graph Databases: Java Developers Navigate the AI Revolution
AI tools keep landing on Java teams faster than anyone can evaluate them. The real question is not whether to use them, but how to keep production code reliable when agents start writing parts of it. For this episode we gathered short interviews from Devoxx and JFall, with Marianne Hoornenborg, Viktor Gamov, Baruch Sadogursky, Stephen Chin, Mario Fusco, Jeroen Benckhuijsen, Martijn Dashorst, Maarten Mulders, and Simon Maple in Foojay Podcast #86.
What we talked about
- Token computation costs and what they mean for AI system design
- A comparative test of six AI coding tools
- Graph databases as a path to more reliable AI responses
- LangChain4J and agent-based task decomposition
- Enterprise Java complexity and the value of deep developer expertise
- Production-ready approaches to AI in Java projects
- Spec-driven coding with AI assistants
What stood out
The conversations keep circling back to the same point. AI does not remove the need for Java skills. It raises the bar, because someone still has to frame the problem, judge the output, and ship code that holds up in production.
See the Foojay Podcast #86 for all info, shownotes, links, etc.