Spring AI Full Course 2026 ☕ | Beginner to Advanced + Production-Ready Java AI App

A focused 18-minute engineering tour of Spring AI for Java developers. We start at the front door — ChatClient, prompts, streaming, structured output — and work outward to advisors, conversation memory, tool calling, embeddings, vector stores, RAG, the Model Context Protocol, observability, testing, and security. You finish with a runnable Spring Boot 3.5 + Java 21 application that calls OpenAI through Spring AI 1.1.x and ties every primitive together. What you get in this course: The shape of Spring AI in one map. Real Java code, in real Maven, with a reference repo you can run. A SupportDesk AI application: a single endpoint that retrieves from a vector store, calls an authorized read-only tool, and replays per-user memory. Production engineering: validation, authorization, observability, cost control, tenant isolation. A reference interface catalogue, glossary, version manifest, and production checklist. Target versions (verbatim in the supplementary notes): Java 21 Spring Boot 3.5.x Spring AI 1.1.x OpenAI (the demo provider); Spring AI 2.0 noted at the end. What the course does not do: It does not promise every Spring AI surface was exhaustively explored in 18 minutes. The supplementary catalogue takes the rest of the surface and points at official docs. It does not claim tests pass against a particular model. The integration test is gated on a real OPENAI_API_KEY and skips without one. It does not promote any single provider as the only choice. Where to learn more: https://docs.spring.io/spring-ai/refe... https://spring.io/projects/spring-ai https://github.com/spring-projects/sp... Chapters are in the chapters file alongside this one. #springai #java #springboot #ai #llm #rag #tutorial