Romain Yon
Member Of Technical Staff at Open Athena
San Diego, California, United States
Summary
🤩
Rockstar🎓
Top SchoolRomain Yon is a seasoned machine learning and infrastructure engineer with 14 years of experience building data-driven systems and ML platforms across startups and large firms. He has deep hands-on expertise in ML infrastructure, recommendations, and cloud-native deployment from senior roles at Spotify and as VP AI Systems Architect at Sotheby's to his current work on open LLMs at Open Athena. A pragmatic engineer and founder, he co-founded Linkurious and Tabtab Labs and spent several years as an indie builder, demonstrating both product intuition and implementation grit. Romain is an active open-source contributor to high-profile projects—improving deployment and Kubernetes integrations for Spotify's Luigi, enhancing data processing in Scio, and adding core functionality and tests to Annoy—bridging research code and production reliability. He combines a formal ML education from Georgia Tech with broad backend, C++/Python, and DevOps skills, frequently focusing on robust error handling and reproducible pipelines. Colocated in San Diego, he repeatedly surfaces non-obvious value by tightening deployment observability and seeding deterministic behavior in probabilistic systems.
14 years of coding experience
11 years of employment as a software developer
Master's Degree, Machine Learning, Master's Degree, Machine Learning at Georgia Institute of TechnologyMaster's Degree, Computer Engineering, Master's Degree, Computer Engineering at Université de Technologie de Compiègne (UTC)
Github Skills (25)
kubernetes10
c-language10
python10
luigi10
scala10
orchestration-framework10
tfrecord10
kubernetes-pods10
data-processing10
cprogramming-language10
devops10
apache-beam10
nearest-neighbors9
testing9
unit-testing9
Programming languages (9)
Github contributions (5)
spotify/luigi
Jul 2014 - Jul 2018
Luigi is a Python module that helps you build complex pipelines of batch jobs. It handles dependency resolution, workflow management, visualization etc. It also comes with Hadoop support built in.
Role in this project:
DevOps Engineer & Cloud Engineer
DevOps Engineer & Cloud EngineerContributions:18 commits, 6 PRs, 24 pushes in 4 years
Contributions summary:Romain primarily contributed to the project's infrastructure and deployment aspects. They modified the Docker runner, implemented and enhanced Kubernetes integration for job scheduling and management, and included instructions for streaming pod logs. The changes involved improving error handling, enhancing the existing job monitoring, and ensuring correct resource management for a more stable deployment pipeline.
pythonbatch-jobsdependency-resolutionorchestration-frameworkworkflow-management
spotify/annoy
Mar 2017 - Nov 2017
Approximate Nearest Neighbors in C++/Python optimized for memory usage and loading/saving to disk
Role in this project:Backend & QA Engineer
Backend & QA EngineerContributions:6 commits, 2 PRs, 1 push in 9 months
Contributions summary:Romain primarily focused on implementing and testing a seeding functionality for the Annoy index, adding the `set_seed` method to the Python and C++ components. They added unit tests to verify the functionality. Their contributions included modifying the C++ code to integrate seeding with the `rand` function and adding tests to assert the seed implementation.
memorymemory-usagepythonapproximate-nearest-neighbor-searchkd-tree
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.