Crafting Intelligent Supersonic Subatomic applications with Quarkus
summary
Quarkus integrates with langchain4j to call LLM APIs (OpenAI / local llama‑style AMA), with a prototype done in six weeks, exposing LLMs as CDI beans and REST clients, supporting conversation memory (default in‑memory size 10), function/tool calls (example: getCurrentTime invoked automatically), Redis vector store for retrieval‑augmented generation of a 100‑movie dataset, and Dev mode prompt iteration plus optional native compilation; demos work end‑to‑end though local models run much slower than OpenAI.
AI-written summary · about
chapters (12)
- 0:00AI and Quarkus journey
- 7:03Large language models
- 10:15L chain for Java
- 12:03Quarkus and L chain
- 14:08Live coding Quarkus AI
- 22:24Conversation state with memory
- 28:56Real-time AI functions
- 35:02Ollama for local AI
- 37:15Enterprise AI features
- 38:21RAG with vector databases
- 41:54AI future and ethics
- 43:01Q&A
related talks
36mCrafting Intelligent Cloud Native applications with Quarkus
43mCrafting Intelligent Applications with Quarkus/LangChain4j
44mUnlocking the Future of Java: LLM Integration Unveiled by Quarkus & LangChain4j
48mAgentic AI with Quarkus, LangChain4j and vLLM
50mDeveloping AI applications in Java with Quarkus and Langchain4j
47mQuarkus and AI Integrating Java and LLM to build intelligent applications
52mQuarkus Supersonic Subatomic Java
01h 01mScaling AI with Java: Building Production-Ready LLM Applications