Create AI Infused Java Apps with LangChain4j with Kevin Dubois Daniel Oh

In their Devnexus presentation, Kevin Dubois and Daniel Oh, both developer advocates, broke down how Java developers can confidently step into the world of AI application development—without leaving the Java ecosystem. While Python has traditionally dominated AI and machine learning, the speakers made a strong case that Java developers already have the tools they need. By using the LangChain4j library alongside the Quarkus framework, developers can seamlessly integrate AI capabilities into their applications, focusing not on training models but on applying them to solve real-world business problems.
The talk walked through the core concepts and practical steps of adding AI functionality to Java projects. With LangChain4j, developers can simplify common challenges such as prompting, handling conversational state with chat memory, and even function/tool calling, which lets an AI model directly execute Java methods. This ability combines the flexibility of probabilistic reasoning with the determinism of existing business logic. Security was also front and center, with a focus on adding guardrails to protect against prompt injection and ensure that models behave predictably and safely within enterprise contexts.
Another highlight was the discussion of Retrieval-Augmented Generation (RAG), presented as a cost-effective way to extend model knowledge without retraining. By injecting custom data—like company policies or documentation—into a model’s context, developers can make AI applications domain-aware. A live demo showed this in action, answering detailed questions about a car rental company’s terms of service, demonstrating how powerful and practical RAG can be for real-world use cases.
The conclusion was clear: Java developers don’t need to abandon their existing expertise to take advantage of AI. With LangChain4j and Quarkus, they have a powerful and familiar toolkit for building the next generation of intelligent, Java-based applications that blend reliability, security, and innovation.
Watch the Session
🎟️ Don’t miss out—watch the full session above, and be sure to join us at Devnexus 2026 for even more insightful talks, live demos, and hands-on guidance on the future of software development.