Summary
Zhenyu Zhou is a Machine Learning Engineer with 10 years of experience building large-scale ML models, GPU inference infrastructure, and document understanding systems, now at Meta after a decade leading ML production at CGG. He combines a PhD-level physics background with deep expertise in distributed systems to design petabyte-scale pipelines, high-throughput GPU clusters, and end-to-end model productization for tens-of-millions document workloads. Notably, he engineered a production ML pipeline handling 50M+ unstructured documents, implemented scalable vLLM-based inference on A100/H100 fleets, and created internal benchmarking and zero-shot labeling frameworks that reduced manual labeling via active sampling. Comfortable working with cross-functional teams and clients, he blends research rigor from his postdoctoral work with practical engineering—shipping reproducible, monitored models with CI/CD and robust orchestration across Redis, Postgres, Neo4j, and Qdrant. Based in Bellevue, WA, Zhenyu’s uncommon mix of condensed-matter physics training and hands-on distributed ML ops enables him to tackle both algorithmic challenges and large-scale system reliability.
10 years of coding experience
17 years of employment as a software developer
Zhenhai high school
Bachelor of Arts (B.A.), Physics, Bachelor of Arts (B.A.), Physics at Peking University
Doctor of Philosophy (Ph.D.), Condensed Matter and Materials Physics, Doctor of Philosophy (Ph.D.), Condensed Matter and Materials Physics at Washington University in St. Louis
English, Chinese