Shang Gao
Applied AI Research Manager at Casetext, Part of Thomson Reuters
Alpharetta, Georgia, United States
Summary
👤
Senior🎓
Top SchoolShang Gao is an Applied AI Research Manager with nine years of experience building scalable deep learning and NLP systems for high-stakes, real-world domains. He currently leads the science team at Casetext/Thomson Reuters developing an agentic legal AI assistant that handles complex multistep legal workflows, automated evaluation of agent outputs, and fidelity of cited sources. His background spans productionizing Transformer-based retrieval and generation pipelines, large-scale annotation programs, and distributed pretraining on leadership-class supercomputers from his work at Oak Ridge National Laboratory. He has driven measurable impact—delivering APIs that reduced human workloads by up to 25% for cancer surveillance—and published practical advances in attention-based models that train far faster than prior state of the art. Comfortable bridging research and engineering, he focuses on robustness, scalability, and reducing hallucinations in generative systems. Based in Alpharetta, GA, he brings a PhD in Data Science and a pragmatic mindset for turning academic innovations into production-ready AI products.
9 years of coding experience
5 years of employment as a software developer
Bachelor of Science (B.S.) Economics, Bachelor of Science (B.S.) Economics at Duke UniversityDoctor of Philosophy - PhD Data Science, Doctor of Philosophy - PhD Data Science at University of Tennessee, Knoxville
Github Skills (15)
machine-learning10
deep-learning10
python10
data-science9
tensorflow9
neural-network9
ml8
deep-neural-networks8
gpu8
numpy5
acceleration3
gpu-acceleration3
tensor3
autograd2
javascript2
Programming languages (2)
Github contributions (5)
iamshang1/Projects
Aug 2016 - Jan 2022
Contributions:189 commits, 1124 pushes, 1 branch in 5 years 5 months
naritapandhe/Microsoft-Malware-Classification-Challenge
Aug 2016 - Sep 2016
Contributions:89 commits in 22 days
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.