Çağlayan Tuna
Research Engineer at Institut Pasteur
Paris, Ile-de-France
Summary
👤
Senior🎓
Top SchoolÇağlayan Tuna is a research engineer with nine years of experience applying machine learning to spatio-temporal and longitudinal data, from satellite-image time series in his PhD to neural approaches for X-ray and behavioral video analysis. He has strong hands-on expertise in tensor methods and contributed to the popular TensorLy library by implementing non-negative CP/Tucker decompositions and HALS updates, bridging academic algorithms with production-ready Python code. His work spans the research–industry interface, including collaborative projects with Airbus Helicopters and Paris brain research centers, where he deployed YOLO and Faster-RCNN for applied detection tasks. Comfortable with high-dimensional, noisy datasets, he combines methodological rigor with pragmatic engineering and an eye for reproducible open-source development. Based in Paris, he now focuses on longitudinal neurological data and deep-learning video analytics, bringing a rare mix of remote-sensing foundations and biomedical applications.
9 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy - PhD, Doctor of Philosophy - PhD at University of South BrittanyMaster's degree, Satellite Communication and Remote Sensing, Master's degree, Satellite Communication and Remote Sensing at Istanbul Technical UniversityKabataş Erkek Lisesi
Turkish, English, French
Github Skills (7)
tensorrt10
machine-learning10
tensor10
python10
numpy10
decomposition10
algebra9
Programming languages (2)
Github contributions (5)
tensorly/tensorly
Jan 2021 - Aug 2022
TensorLy: Tensor Learning in Python.
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:46 reviews, 134 commits, 34 PRs in 1 year 7 months
Contributions summary:Çağlayan contributed to the development of tensor decomposition methods, including non-negative CP decomposition, and the implementation of HALS (Hierarchical Alternating Least Squares) algorithm for non-negative Tucker decomposition. They added the function of non_negative_tucker_hals function with several updates related to core update methods and sparsity coefficients. They also worked on the implementation of smoothness and monotonicity proximal operators.
pythontensortensorlymachine-learningdecomposition
caglayantuna/tensorly
Jan 2021 - Sep 2022
TensorLy: Tensor Learning in Python.
Contributions:2 PRs, 341 pushes, 95 branches in 1 year 8 months
pythontensortensorly
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.