Hugo Yeche
Research Scientist at REOR20
Zurich, Zurich, Switzerland
Summary
👤
Senior🎓
Top SchoolHugo Yeche is a research scientist and ML PhD with eight years’ experience building temporal deep learning models for healthcare and environmental forecasting. Trained at ETH Zurich’s Rätsch lab, his thesis and postdoctoral work translated event prediction models for ICU time series into clinically deployable early warning systems. He contributes to open-source ML tooling—having improved the popular keras-tcn implementation by refining residual block architectures and ensuring fidelity to the original TCN paper. Now based in Zurich and working on real-time flood forecasting, he blends rigorous academic research with hands-on engineering to deliver robust, production-ready models. Notably, his background spans both applied clinical trials and large-scale field deployments, giving him rare end-to-end experience from algorithm design to real-world impact.
8 years of coding experience
2 years of employment as a software developer
Master 2 Applied Mathematics, Master 2 Applied Mathematics at ENS Paris-SaclayCPGE Blaise PascalTélécom ParisMaster of Science - MS Data Science, Master of Science - MS Data Science at EURECOMDoctor of Philosophy - PhD Machine Learning, Doctor of Philosophy - PhD Machine Learning at ETH ZürichLycee Blaise Pascal
English, French
Github Skills (7)
keras10
machine-learning10
recurrent-neural-networks10
deep-learning10
python9
neural-network9
convolutional-neural-networks9
Programming languages (4)
Github contributions (5)
philipperemy/keras-tcn
Apr 2020 - Apr 2020
Keras Temporal Convolutional Network.
Role in this project:
ML Engineer
ML EngineerContributions:5 commits, 1 PR, 2 comments in 1 day
Contributions summary:Hugo primarily worked on modifying and refining the Temporal Convolutional Network (TCN) implementation within the `keras-tcn` repository. Their commits focused on adjusting the architecture of the residual blocks, including conditional 1x1 convolutions, removing and reintroducing convolutions, and making the parameters of the model match those described in the original research paper. They also made changes to the build process. The changes suggest an effort to improve the model's architecture and functionality.
temporal-convolutional-networkdeep-learningrecurrent-neural-networksconvolutionalmachine-learning
ratschlab/HIRID-ICU-Benchmark
Jun 2021 - Sep 2022
Repository for the HiRID ICU Benchmark (HiB) project
Contributions:2 reviews, 66 commits, 14 PRs in 1 year 3 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.