Fernando Pérez-García
Cambridge, England, United Kingdom
Summary
🤩
RockstarFernando Pérez-garcía is a Senior AI Research Engineer with 10 years’ experience applying AI and image processing to improve epilepsy treatment, currently at Microsoft Research Cambridge while completing a PhD at University College London and King’s College London. He blends deep domain expertise in medical imaging—segmentation, electrode localization and loss-function design—with production ML engineering, notably refactoring the InnerEye inference pipeline and integrating TorchIO for robust patch-based inference. His contributions span open-source neuroimaging projects (nibabel, nipype), clinical tooling (3D Slicer homebrew cask), and platform integrations (Microsoft OAuth for Girder), showing an ability to move research into clinical-ready systems. Based in Cambridge, he pairs rigorous research instincts with pragmatic software craftsmanship, often improving documentation and reproducibility alongside core algorithmic work.
10 years of coding experience
Stackoverflow
635reputation
109kreached
18answers
3questions
Github Skills (61)
pytorch10
javascript10
loss-functions10
oauth210
documentations10
python10
back-end-development10
api-design10
authentication10
machine-learning10
user-authentication10
homebrew-cask10
microsoft-azure10
homebrew10
tensorflow210
Programming languages (16)
Github contributions (5)
microsoft/InnerEye-DeepLearning
Feb 2022 - Aug 2022
Medical Imaging Deep Learning library to train and deploy 3D segmentation models on Azure Machine Learning
Role in this project:
ML Engineer
ML EngineerContributions:83 reviews, 21 commits, 38 PRs in 5 months
Contributions summary:Fernando primarily contributed to the improvement and maintenance of the medical imaging deep learning library. Their work focused on refactoring the inference pipeline, replacing the RadIO library with TorchIO for patch-based inference. They also added features like padding model outputs and ensuring shape compatibility for inputs, and fixed issues with the logging system. Additionally, the user addressed documentation issues and fixed build warnings.
deep-learningmachine-learningazure-machine-learningdeep-learning-libraryimaging
Homebrew/homebrew-cask-versions
Aug 2017 - Nov 2021
💀 Alternate versions of Casks (deprecated)
Role in this project:Full-stack Developer
Full-stack DeveloperContributions:79 commits, 78 PRs, 9 comments in 4 years 3 months
Contributions summary:Fernando primarily focused on updating and maintaining the 'slicer-nightly' cask within the homebrew-cask-versions repository. Their contributions involved updating the version and SHA256 checksum of the cask to reflect new nightly builds of 3D Slicer. They also updated the URL for the download and, in a later commit, refactored the cask definition to use the `:latest` version with a dynamic URL, and included zap stanza for cleaner removal.
caskscaskhomebrewalternate
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.