· Updated
JetBrains launched Koog for Java, bringing its enterprise AI agent framework to Java teams with a fully idiomatic Java API. Java developers can now build fault-tolerant, observable AI agents directly inside existing Spring Boot backends — no Python microservice required.
@Tool and @LLMDescription to expose them to LLMs, then wire agents using fluent builders. The framework supports Spring Boot, all major LLM providers — OpenAI, Anthropic, Google, DeepSeek, Ollama — plus multiple workflow strategies.Originally built as JetBrains scaled its own AI activities, Koog replaces ad hoc prompting with structured, fault-tolerant agent workflows. The Java release means enterprise teams with existing Java backends no longer need Kotlin or a separate Python microservice to run production-grade agents. Agent persistence lets failed runs recover without repeating expensive LLM calls.
Point your Java teams at Koog — the Spring Boot integration and @Tool annotation pattern mean agents drop into your existing codebase alongside your current service layer.
If you build enterprise Java backends, adding AI agents shouldn’t mean spinning up a separate Python service. Koog for Java, an open-source, enterprise-ready AI agent framework for the JVM, has just been released. Learn more: https://t.co/kshql9d3rq
Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →
More like this
CursorMar 4
IntelliJ IDEA, a JetBrains IDEMay 16
Kotlin by JetBrainsJun 4
Elastic DevMar 18