Shuai Zheng
Co-Founder & CEO at Stealth Startup
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolShuai Zheng is a founder-led AI engineer and entrepreneur with nine years of experience building large-scale ML systems, currently leading a stealth AI startup after co-founding Boson.ai to commercialize multi-modal foundation models and enterprise LLMs. He previously drove LLM scaling and distributed training at AWS, optimizing infrastructures across 10k–20k GPU clusters and enabling products like Bedrock, CodeWhisperer, Titan, and SageMaker Model Parallelism. His hands-on ML engineering roots include significant open-source contributions to GluonNLP and MXNet—where he improved training examples, optimizers (adding Adamax epsilon and a LANS optimizer), and test quality—reflecting a blend of production ML, systems optimization, and research. Shuai holds a PhD in Computer Science from HKUST and combines deep academic training with practical achievements like clocking extremely fast BERT-large training on cloud and building on-prem GPU/storage stacks for pretraining and serving. He is equally comfortable leading cross-functional science teams and diving into test automation, linting, and optimizer internals, a detail that underscores his end-to-end craftsmanship in ML systems.
8 years of coding experience
6 years of employment as a software developer
Bachelor's degree Computer Software Engineering, Bachelor's degree Computer Software Engineering at Beijing Jiaotong UniversityHong Kong University of Science and Technology (HKUST)
English, Chinese
Github Skills (14)
mxnet10
optimizer10
nlp10
deep-learning10
gluon10
trainings10
python10
optimizers10
modeling10
test-automation10
testing10
pytest9
software-quality9
machine-learning9
Programming languages (6)
Github contributions (5)
dmlc/gluon-nlp
Apr 2018 - Sep 2020
NLP made easy
Role in this project:
QA Engineer / Test Automation Engineer
QA Engineer / Test Automation EngineerContributions:9 reviews, 47 commits, 68 PRs in 2 years 4 months
Contributions summary:Shuai primarily focused on improving the quality of the NLP toolkit by rearranging and modifying test structures, fixing linting issues, and ensuring the correct execution of tests. They corrected test setups and implementations within the `scripts/nmt` and `scripts/` test hierarchy. Their contributions included adjustments to test configurations to ensure correctness and address issues within existing test cases. The user also addressed code issues to improve code quality and facilitate more comprehensive and reliable testing.
nlpmxnetnatural-language-understandingnlgnatural-language-generation
apache/mxnet
Apr 2018 - Jun 2020
Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
Role in this project:ML Engineer
ML EngineerContributions:5 reviews, 7 commits, 8 PRs in 2 years 3 months
Contributions summary:Shuai primarily contributed to the MXNet repository by fixing and improving examples related to Gluon, a high-level API for deep learning. Their work included correcting language model examples, modifying training scripts, and updating documentation files. Furthermore, the user refactored optimizer code and added an epsilon parameter to the Adamax optimizer, suggesting a focus on model training and optimization techniques. The user also added a new LANS optimizer, indicative of contributions to model training methods within the MXNet ecosystem.
pythonschedulerdataflowmutationdata-science
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.