Summary
Chengyan Fu is a software engineer with 13 years of experience and over eight years leading the design and production rollout of hyperscale storage and compute infrastructures. He has a proven track record of building teams and zero-to-one systems—ranging from multi-region distributed storage and cost-efficient hardware tiers (SMR, tape) to modern pretraining and speculative decoding pipelines for large language models. Currently at Meta working on MSL infra and PyTorch pre-training infrastructure, he combines deep systems engineering with applied ML model-architecture exploration. Prior roles at Databricks and multiple stints at Meta/Facebook show strong domain fluency in lake storage, file formats, and data governance for training and inference. Based in the Greater Seattle Area, he brings a pragmatic bent for piloting new technologies and shipping resilient solutions at scale. A subtle differentiator is his continuity across both hardware-aware storage optimization and cutting-edge ML infra, enabling cross-layer improvements that few engineers span end-to-end.
12 years of coding experience
11 years of employment as a software developer
Master's degree Computer Software Engineering, Master's degree Computer Software Engineering at Shanghai Jiao Tong University