Alexander Terenin
Assistant Research Professor at Cornell University
City of Ithaca, New York, United States
Summary
👤
Senior🎓
Top SchoolAlexander Terenin is an Assistant Research Professor at Cornell with a decade of experience in AI and algorithmic decision-making, blending rigorous academic training (PhD from Imperial College, postdoc at Cambridge) with industry research roles at Petuum and internships at Microsoft and eBay. He focuses on probabilistic and scalable machine learning methods and has hands-on contributions to ML tooling, including enhancing Flux.jl by making Cholesky decompositions autograd-friendly and GPU-tested. Based in Ithaca, he bridges theory and practice—publishing and teaching while shipping robust code for real-world ML stacks. Colleagues know him for careful numerical work and ensuring that sophisticated linear-algebra primitives remain usable and well-tested in modern differentiable frameworks.
10 years of coding experience
2 years of employment as a software developer
Postdoc, Machine Learning, Postdoc, Machine Learning at University of CambridgePhD, Statistics (Machine Learning), PhD, Statistics (Machine Learning) at Imperial College LondonB.S. Statistical Science and B.A. Psychology, B.S. Statistical Science and B.A. Psychology at UC Santa BarbaraUniversity of California Santa Cruz
Github Skills (11)
neural-network10
machine-learning10
deeplearning-ai10
fluxor10
deep-learning10
artificial-neural-networks10
julia10
flux10
data-science7
cuda7
testing6
Programming languages (18)
Github contributions (5)
FluxML/Flux.jl
Apr 2020 - Jul 2020
Relax! Flux is the ML library that doesn't make you tensor
Role in this project:
ML Engineer
ML EngineerContributions:5 commits, 7 PRs, 28 comments in 2 months
Contributions summary:Alexander focused on enhancing the Flux.jl machine learning library, specifically by adding and testing functionality related to the `Cholesky` decomposition within the framework. They introduced a functor for `Cholesky`, enabling its use within the autograding system. Further, the user added a test case to ensure that `Cholesky` works correctly on the GPU, demonstrating their commitment to ensuring the library's broad applicability and testing its CUDA support. Finally, the user made `Cholesky` non-trainable.
ml-librarythe-human-braindata-sciencedeep-learningneural-networks
aterenin/phdthesis
Jun 2021 - Apr 2022
Contributions:8 releases, 139 commits, 163 pushes in 10 months
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.