From CLIs to LLMs: 5 Cool Applications to Build with GraalVM
summary
Five concrete builds: a tiny CLI using Spring Shell (Picocli recommended) that starts in ~58 ms; a full‑stack Vaadin travel recommender using Spring AI and OpenAI that launches <100 ms as an ~180 MB native binary; a pure‑Java LLM inference engine (llama3.java) producing a ~66 MB native assistant that boots in ~7 ms and runs on CPU; benchmarking on Spring PetClinic shows JVM C2 reaches ~1.5k RPS after warmup, plain native ~1.2k RPS, PGO + G1 optimized native reaches ~1.5k RPS; native images cut energy usage by ~33%, showing native image and LLM inference work in practice.
AI-written summary · about
chapters (9)
- 0:00Introduction to GraalVM
- 1:59What is GraalVM
- 6:57CLI app with Spring Shell
- 10:30Full-stack web app demo
- 16:11Local LLM in pure Java
- 22:31Benchmarking JVM vs native
- 36:02Security and SBOM in native
- 42:11Future GraalVM projects
- 46:24Resources and closing
related talks
48mFrom CLIs to LLMs: 5 cool applications to build with GraalVM
50mGoing AOT: Everything you need to know about GraalVM for Java
53mBootiful GraalVM
47mBring the action: using GraalVM in production
35mTop 10 Things to Do with GraalVM
46mBring the Action: Using GraalVM in Production
47mGoing Native: Fast and Lightweight Spring Boot Applications with GraalVM
46mFast and Lightweight Spring Boot Applications with GraalVM