About this event
JUG Noord brings together developers in the northern Netherlands to explore Java, JVM languages, and modern AI-enabled development practices. The February 2026 meetup centers on pairing humans with AI to build smarter, more reliable software, with a concrete focus on eval-driven development for agentic systems. Expect practical talks, hands-on discussion, and a chance to connect with fellow JVM enthusiasts who are curious about how AI can augment traditional software engineering.
What to expect:
- In-person evening in Groningen (venue opens 16:30; talks begin at 17:00) with food and drinks between sessions
- Two deep-dive talks focused on AI-assisted development on the JVM and how to make agentic software more reliable
- Real-world, practitioner-led content designed for Java and Kotlin developers who want to bring AI into production responsibly
- Emphasis on learning by building and evaluating agents in a deterministic, test-driven style adapted for probabilistic AI systems
Agenda highlights:
- How to build your own fun and absurd pair programmer by Alexander Chatzizacharias (JDriven): a hands-on session exploring building a playful, personality-rich AI assistant using LLMs, Spring Boot, and vector databases. Attendees will dive into Retrieval-Augmented Generation, multi-vector search, and Model Context Protocol integrations, gaining practical insights for creating engaging agentic workflows and personality-rich digital assistants. The talk will illuminate how to design and test interactive AI helpers that collaborate with humans rather than simply automate steps.
- Eval-Driven Development: The Fine Line Between Agentic Success and Failure by Urs Peter (Xebia): a deep dive into the probabilistic nature of agentic systems and the risks of hallucinations, context drift, and evolving multi-step workflows. The session introduces Eval-Driven Development (EDD) as an engineering-first approach—extending test-driven development into agentic architectures. You’ll learn how to test agents at multiple layers (schema validation, tool correctness, decision flows, end-to-end goals), gather meaningful metrics, turn agent traces into actionable insights, and construct a continuous evaluation loop using real-world data so agents improve over time. The talk showcases JVM AI frameworks (Spring-AI, LangChain4j, Koog) in practical demonstrations.
Who should attend:
- Java and Kotlin developers curious about injecting AI into JVM applications
- Engineers exploring agentic wo
