Name:Yu Kang

Bio:

Yu Kang is a Senior Expert / Research Scientist at the Software Engineering Application Technology Lab, Huawei. His research lies at the intersection of AI, software engineering, and systems. It centers on agent evolution and recursive self-improvement, LLM training and agentic reinforcement learning, and product-grounded environments, data, and evaluation for coding agents. A distinctive thread of his work is modeling the complexity of real products as scientific problems. He pioneered turning industrial product repositories and product tasks into executable data and environments for evaluating, tuning, and training coding agents.

Before joining Huawei, he was a Principal Research Manager at Microsoft DKI (Data, Knowledge, Intelligence; formerly part of Microsoft Research Asia) from 2018 to 2026. There he led SWE-bench-Live (NeurIPS 2025) and RepoLaunch, which Zhipu AI used to build executable training environments for GLM-5, and contributed to Agent Lightning, UFO, and TaskWeaver. His AIOps and resource-management research was integrated into Microsoft’s fundamental cloud services behind Azure and Microsoft 365. His 60+ papers include an ACM SIGSOFT Distinguished Paper Award (ESEC/FSE 2022) and a Best Research Paper Award (ISSRE 2021). He is an adjunct master’s supervisor at the School of Computer Science, Fudan University, and received his PhD from The Chinese University of Hong Kong.

Country:China

Affiliation:Huawei

Research interests:Software Engineering, AI, Cloud computing