Scala Meets GenAI - Build the Cool Stuff with LLM4S
summary
LLM4S is a Scala batteries‑included toolkit that in six months goes from empty repo to a usable stack: basic LLM calls across OpenAI/Anthropic, RAG over custom datasets using vector stores (Postgres/Chroma/Azure), an agentic loop with tool calling and sandboxed code execution, and multimodal pipelines for image + music + speech. Demo app is a procedurally generated Zork‑style game (live demo falters but re-recording shows features), GSOC adds 4 major contributions, and tracing integrates Langfuse.
AI-written summary · about
chapters (10)
- 0:00Scala and GenAI intro
- 1:04Why Scala for GenAI
- 2:07Zork demo with LLM4S
- 15:32Peeling back the layers of LLM4S features
- 19:50Modern LLM workflows
- 23:22GSoC student contributions
- 29:04Multimodal and tracing in LLM4S
- 33:26Scala for AI excellence
- 34:42Scala and AI community
- 39:15Q&A
related talks
01h 34mA Practical Starter's Guide to building LLM based projects
46mScala on the Move!
01h 01mScaling AI with Java: Building Production-Ready LLM Applications
34mConcerto for Java & AI – Building Production-Ready LLM Applications
02h 49mJava meets AI: Build LLM-Powered Apps with LangChain4j
02mSCALAR 2023
27mBuilding Useful Apps With Small, Local LLMs
27mNavigating Scala’s AI Landscape