Agentic AI evolves from traditional AI by autonomously deciding which external tools and resources to use, unlike systems with predefined workflows.
While Python has many agentic AI frameworks, Java developers can combine tools like Quarkus, Spring AI, and LangChain4j to build similar applications.
Use the Quarkus application generator to quickly create a new project with dependencies for OpenAI and MCP, then run it in development mode for live coding.
A simple Java interface with LangChain4j annotations and a carefully crafted system prompt can instruct the AI model to autonomously utilize any available tools.
Configure external tools like search, maps, and messaging by defining MCP servers in the application properties file, which Quarkus automatically downloads and runs.
A live demonstration shows the agent autonomously using search, maps, and Slack tools to fulfill a complex user request for finding and sharing restaurant recommendations.
Simplify multi-agent development and collaboration by using an Internal Developer Platform (IDP) like Backstage, with Quarkus automatically generating software templates from your project.
The MCP roadmap includes a server registry and multi-modal support, and the key takeaway is that Quarkus simplifies complex agentic AI development for Java developers.
msg
Ismaning, Germany
29:18Daniel Oh & Kevin Dubois
30:01Mario Fusco
39:11Thomas Dohmke
32:02Kevin Dubois
28:21Daniel Oh
41:17Syed M Shaaf
27:43Anu Bharadwaj & Tobias Schlottke
25:14Ivan Burazin
Jobs that call for the skills explored in this talk.
Siemens AG
Bellevue, United States of America
autonomous-teaming
München, Germany
fleetbay GmbH
Koblenz, Germany
Agile Talent
Tata Consultancy Services Limited
Jacksonville, United States of America
