Wayne Xiong
Managing Director at WorldQuant
New York City Metropolitan Area United States
Summary
🤩
Rockstar🎓
Top SchoolWayne Xiong is a seasoned technology leader who has spent the past decade building and operating large-scale Language AI systems, most recently leading Microsoft’s Language AI organization and GenAI training infrastructure before moving to WorldQuant as Managing Director. He manages global teams of 100+ researchers, applied scientists, and engineers to deliver high-quality, low-latency services—processing trillions of tokens daily across translation, dialog, summarization, and extraction use cases—and also architected unified pipelines for data curation, training, fine-tuning, distillation and compliance. Wayne’s career blends foundational research (including leading the 2016 effort that achieved human parity in conversational speech recognition) with hands-on engineering and production-grade ML infrastructure work, including contributions to the Microsoft Cognitive Toolkit (CNTK) to enable multi-GPU evaluation. He’s known for translating breakthroughs into practical, compliant enterprise deployments and for innovating at the intersection of model quality, scalability, and reliability.
10 years of coding experience
15 years of employment as a software developer
B. Eng, Electronic Engineering, B. Eng, Electronic Engineering at Tsinghua UniversityThe Chinese University of Hong Kong (CUHK)
Github Skills (8)
machine-learning10
c-language10
cntk10
deep-learning10
cprogramming-language10
dotnet-core4
python4
csharp4
Programming languages (2)
Github contributions (5)
microsoft/CNTK
Feb 2016 - Aug 2016
Microsoft Cognitive Toolkit (CNTK), an open source deep-learning toolkit
Role in this project:
ML Engineer
ML EngineerContributions:20 commits, 233 pushes, 21 branches in 6 months
Contributions summary:Wayne's contributions primarily focused on enhancing the cross-validation capabilities within the Microsoft Cognitive Toolkit (CNTK). They modified the SimpleEvaluator class to support multi-GPU processing, optimizing deep learning model evaluation. Further work involved adjusting model evaluation before and after training epochs and applying code review feedback, indicating a focus on code quality and performance improvements within the deep learning framework. They also added an end-to-end test case.
pytorchpythondeep-learningc-plus-plusmachine-learning
Azure-Samples/Cognitive-Services-Speech-Devices-SDK
May 2018 - May 2018
The sample app and documentation of the Microsoft Speech Devices SDK.
Contributions:7 pushes in 3 days
azure-cognitive-servicesspeech-recognitionsdkcognitive-servicesmicrosoft-cognitive-services
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.