ServingStudio uses measured simulation, detailed performance analysis, and an agent to improve LLM serving systems. It compares configurations using performance models grounded in real GPU measurements, then lets an agent build and validate the strongest candidates in real serving frameworks.
Visit the project website, read the blog, or get started on GitHub.
Kan Zhu 朱侃
PhD Student (Computer Science)
I am Kan Zhu, a fourth year PhD student at University of Washington’s Paul G. Allen School of Computer Science and Engineering.