Workshop: Bringing AI to Java: from Function Calls to Context-Aware Applications

22 Jan 2026

We are looking forward to a workshop by Andras Fejes about building AI applications using Quarkus and LangChain4J.

Speaker: Andras Fejes Language: English Location: Technische Hochschule Mannheim

Anmeldung

Please register on our Meetup page to help us with planing. As this is a workshop with limited space, please also unregister if you can’t make it.

Location

For the first time, the Technische Hochschule Mannheim will be our host. Thanks a lot for this opportunity! We also welcome all interested students!

The workshop will take place in the Computer Science (Informatik) building (Speyerer Str. 4, 68163 Mannheim) in room A108. Please note that parking is very limited, so rather take the tram line 1 to tram stop “Hochschule”. Enter from Speyerer Str. 4, either from the street (you need to find your way around the fence of the building site) or via the entry to the parking space within the campus on the left of the building and entering the building from the rear.

Agenda

• 18:30 Uhr: door is open
• 19:00 Uhr: start of the workshop
• 20:30 Uhr: end of the workshop

Abstract

AI is rapidly moving from experimental prototypes to reliable components of modern software systems. For Java developers, the challenge is no longer if AI can be integrated, but how to do it effectively—bridging the gap between traditional enterprise architecture and the dynamic, context-driven world of Large Language Models (LLMs).

This session explores how LLMs and orchestration frameworks like LangChain4J can be leveraged in Java applications using function calling, tool integration, and the Model Context Protocol (MCP). We’ll look at how these concepts transform AI from a conversational assistant into a predictable, composable system component.

The discussion will cover:

The principles of LLM function calling and how context transforms AI responses from “guesses” into deterministic operations.

How to integrate AI workflows into Java backends using Quarkus and modern APIs.

The trade-offs, pitfalls, and real-world lessons learned while building production-ready AI-assisted tooling.

This is a workshop-style session designed for both hands-on participants and observers. If you plan to explore the code examples on your own machine, please ensure your laptop has at least 32 GB of RAM, as AI workloads are not lightweight.

Join us to discover how AI can become a practical and dependable part of Java-based systems—empowering developers to build applications that are not just intelligent, but contextually aware and operationally sound.

Bio

With over 25 years of experience in IT, Andras Fejes has worked across diverse business domains, including finance, public transport, telecommunications, media, bioinformatics, and more. He is highly skilled in Java technologies, with deep expertise in Spring Boot, WildFly, Quarkus, and Java/Jakarta EE. Currently, Andras is a Software Engineer at LivePerson, where he focuses on building real-time, cloud-based services using Spring Boot, Kafka, and Cassandra. He’s passionate about creating high-performance, scalable solutions and exploring the latest innovations in cloud-native development.

Eine Veranstaltung des iJUG e.V., organisiert durch die JUG Mannheim (majug).