Yusuke Niitani
Software Engineer at Preferred Networks, Inc.
Tokyo, Japan
Summary
🤩
Rockstar🎓
Top SchoolYusuke Niitani is a software engineer based in Tokyo with 11 years of experience building and optimizing ML and scientific computing libraries. At Preferred Networks he has focused on backend, MLOps and performance-sensitive work, contributing substantial refactors and integration tests to ChainerCV and core array operations in CuPy and Chainer (including advanced indexing, broadcast/concatenate and scatter_add). His contributions show a blend of practical production engineering and deep numerical/array expertise, plus attention to edge cases and test coverage. He also has experience in scientific visualization (glumpy) and a research profile visible on Google Scholar, reflecting a strong academic grounding from the University of Tokyo. Notably, his work often bridges model-format conversion, low-level array semantics, and deployable ML tooling rather than just research prototypes.
11 years of coding experience
2 years of employment as a software developer
University of Tokyo
Stackoverflow
286reputation
69kreached
2answers
0questions
polygon
top-5%
Github Skills (31)
python10
image-processing10
chainer10
rendering10
data-manipulation10
machine-learning10
render10
numpy10
manipulation10
gpu10
cupy10
cuda10
model-conversion10
opengl10
linear-algebra10
Programming languages (7)
Github contributions (5)
chainer/chainercv
Feb 2017 - Dec 2019
ChainerCV: a Library for Deep Learning in Computer Vision
Role in this project:
Back-end Developer & MLOps Engineer
Back-end Developer & MLOps EngineerContributions:14 releases, 2217 commits, 757 PRs in 2 years 10 months
Contributions summary:Yusuke's commits primarily focus on refactoring the existing Python code, especially for the implementation of image manipulation in a research library, and also on converting models trained in a variety of formats to chainer format. They modified the base library and some model interfaces and also implemented integration tests of the code they work on.
deep-learningpytorchcomputer-visionvision
chainer/chainer
Nov 2016 - Jul 2019
A flexible framework of neural networks for deep learning
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:330 commits, 77 PRs, 182 comments in 2 years 8 months
Contributions summary:Yusuke primarily contributed to the CuPy library, focusing on core array operations and enhancing advanced indexing capabilities. They implemented and optimized functions such as `broadcast_to` and `concatenate`, and introduced a new `scatter_add` function. Furthermore, they added tests for the implemented functions, showcasing a focus on improving functionality, numerical stability and performance.
cudapythonmxnetcaffe2flexible-framework
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.