Panel Discussion: LangChain4j, a year later.
summary
LangChain4j now has over 150 contributors and ships concrete features: Gemini multimodal support with a 2 million token context and PDF/audio/video handling, Quarkus integration that injects observability & telemetry spans and metrics for real production traces, a BM25 re-ranker prototype that “just worked” in Java, CFP anonymization and semantic search in production, guardrails/moderation checks, and prompt-caching plus a 1.0 roadmap underway.
AI-written summary · about
chapters (11)
- 0:00Introduction to panel
- 2:11Favorite features in LangChain4j
- 8:37Gemini’s impact on Java AI
- 11:07Azure AI integration insights
- 12:51Stefan’s LangChain4j projects
- 16:22Quarkus LangChain4j origins
- 20:00LangChain4j developments past year
- 25:02AI’s future in Java ecosystem
- 37:00Guardrails and safety in AI systems
- 45:35AI skills for software engineers
- 51:12Closing and call to action
related talks
53mJava meets AI: Build LLM-Powered Apps with LangChain4j
01h 01mCreate #AI Infused #Java Apps with #LangChain4j
46mCodepocalypse Now Lang Chain4j vs Spring Al
50mDeveloping AI applications in Java with Quarkus and Langchain4j
49mCodepocalypse Now: LangChain4j vs. Spring AI
47mJava Meets AI: Integrating LLMs into Your Applications with LangChain4j
37mTaming the Information Chaos: Building an AI Ally with Langchain4J
45mBuilding Intelligent Enterprises: AI-Powered Java with LangChain4j & CDI