Summary
Yuchen Tian is a Senior Research Engineer with 8 years of experience building ML-driven systems and developer tools at AWS, focused on making complex problems tractable and useful in daily workflows. At Amazon he designed SWEBench to measure LLM agentic coding capabilities, built large-scale RL environment pipelines, and contributed to CodeWhisperer’s online code generation, evaluation, debugging, and customization. With an MEng from Carnegie Mellon and a background in search, relevance, and forecasting, he blends production engineering rigor with research-led experimentation. He’s also an active open-source practitioner—authoring zi2zi, an automatic typeface generation project—reflecting an interest in creative ML applications beyond core infrastructure. Colleagues describe him as a builder who “lives in the future, then builds what is missing,” bringing pragmatic creativity to large-scale AI tooling.
8 years of coding experience