LLM Microserving: a new RISC-style approach to design LLM serving API (blog.mlc.ai)

Scale LLM serving with programmable cross-engine serving patterns, all in a few lines of Python