Zhucheng Tu
Research Scientist Engineer at Luma AI
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolZhucheng Tu is a Research Scientist/Engineer based in San Francisco with 13 years of experience building and productionizing large-scale ML and multimodal systems. Currently working on multimodal models at Luma AI, he previously led the PyTorch implementation of a foundation model for personalized recommendations at Netflix, owning pretraining, fine-tuning, batch inference, and homepage deployments. At Apple he built search and open-domain QA systems and was a top application-team contributor to axlearn, bridging JAX research frameworks with production needs. He combines deep research experience (MMath Computer Science, Waterloo) with hands-on engineering skills in mixed-precision training, muP, and kernel integrations that materially improved training efficiency. Known for turning research prototypes into robust infrastructure, he has a track record across product-critical ML systems spanning recommendation, QA, and privacy-aware assistant components. Outside work he maintains an active public presence—code, papers, and commentary—making him fluent in both research and engineering cultures.
13 years of coding experience
8 years of employment as a software developer
International Baccalaureate Diploma, International Baccalaureate Diploma at Western Canada High SchoolMMath Computer Science, MMath Computer Science at University of Waterloo
English, Chinese
Github Skills (149)
deep-learning10
text-processing10
streaming10
markdown10
convolutional-neural-networks10
batch10
lucene10
retrieval10
keyword10
information-retrieval10
semantic-textual-similarity10
java10
static-site10
graphene10
liquid10
Programming languages (11)
Github contributions (5)
apple/ml-qrecc
Oct 2020 - May 2022
Contributions:11 commits, 10 pushes, 8 comments in 1 year 7 months
tuzhucheng/Anserini
Mar 2017 - Jul 2018
An information retrieval toolkit built on Lucene
Contributions:44 pushes, 15 branches, 7 comments in 1 year 4 months
retrievalluceneinformation-retrieval
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.