Yusong Gao

I build production infrastructure for large-scale data and AI systems.

Today, I focus on large-scale LLM inference at Ant Group, including serving architecture, scheduling, performance, reliability, and observability.

Previously, I built production systems for PolarDB and Alibaba Cloud RDS.

• Large-Scale AI Inference — serving architecture, scheduling, performance, reliability, and observability.
• Cloud Database Infrastructure — production systems for PolarDB and Alibaba Cloud RDS.
• Open Source — work around SGLang and llm.cpp.