Marc Romeyn is a Senior Machine Learning Engineer with 11 years of experience building production-grade recommendation and large-scale deep learning systems, currently contributing to NVIDIA Merlin and Transformers4Rec for GPU-accelerated recsys. Previously a core contributor to Spotify's centralized ML platform and an Entrepreneur First alum who co-founded Manner.ai, he combines product-minded engineering with research-driven modeling. His work spans PyTorch/TensorFlow integrations, sequence masking and data augmentation for recommendations, and engineering at scale using Spark, Airflow and Kubernetes. Based in Stockholm, he bridges academia and industry with master's training in distributed and data-intensive systems and a practical knack for refactoring complex ML codebases to improve maintainability and performance.
11 years of coding experience
9 years of employment as a software developer
Master’s Degree, EIT Digital Master School - Data Intensive Computing, B, Master’s Degree, EIT Digital Master School - Data Intensive Computing, B at KTH Royal Institute of Technology
Master’s Degree, EIT Digital Master School - Distributed Systems and Services, 2.0, Master’s Degree, EIT Digital Master School - Distributed Systems and Services, 2.0 at Technische Universität Berlin
High School, VWO - Natuur & Techniek, High School, VWO - Natuur & Techniek at Berger Scholen Gemeenschap
EIT Master School
Bachelor’s Degree, Knowledge Engineering, Bachelor’s Degree, Knowledge Engineering at Maastricht University
Transformers4Rec is a flexible and efficient library for sequential and session-based recommendation and works with PyTorch.
Role in this project:
Back-end Developer & ML Engineer
Contributions:34 reviews, 174 commits, 29 PRs in 4 months
Contributions summary:Marc primarily focused on refactoring and porting code related to masking sequences and data augmentation within the `transformers4rec` library. Their contributions involved renaming files, restructuring modules, and integrating new features like StochasticSwapNoise. Additionally, the user worked on core aspects, including the implementation of tabular features, model configurations, and improvements to the API. This suggests a focus on expanding the capabilities and maintainability of the library for recommendation tasks, with an emphasis on PyTorch-based functionality.
Merlin Models is a collection of deep learning recommender system model reference implementations
Contributions:301 reviews, 458 commits, 281 PRs in 1 year 6 months
pytorchimplementationsrecommenderdeep-learninggpu
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.
Request Free Trial
Marc Romeyn - Senior Machine Learning Engineer at NVIDIA