This talk will explore the main architectural blueprints for Java GenAI Applications, offering insights into addressing
the most typical requirements for such applications, approaching recurrent problems, and breaking them into manageable components to unlock the potential of LLMs with Java.

The talk comprises the most common design patterns for building GenAI solutions with Java, and it will provide some reference architectures and solutions for building LLM-based agents. Code examples with Spring AI, LangChain4J and OpenAI will be presented, along with an Oracle Database to support the examples.