Summary
Marc Van Oudheusden is a Staff Machine Learning Engineer with 11 years of focused ML/AI experience and a two-decade engineering background spanning finance, telecom and startups. He has delivered 20+ bespoke ML solutions—from bond trading recommenders and fraud detectors to medical language models and sports video pipelines—balancing cutting-edge research with pragmatic, production-ready design. At AWS he built probabilistic and graph-based models, temporal link-prediction methods, and generative-document demos; earlier roles at SGCIB combined data science leadership with large-scale platform and cost-optimization wins. Comfortable leading teams of 18+ and cross-functional programs of 30, he pairs strong technical depth (GNNs, I3D/SlowFast, ClinicalBERT, PyTorch/TensorFlow) with product-minded tradeoffs. Trained at École Polytechnique and ENSTA, he brings rigorous applied-maths and operations-research thinking that surfaces in practical innovations like faster risk analytics runtimes and optimized testing algorithms. Based in Nouvelle-Aquitaine, he quietly blends research instincts with delivery discipline—favoring solutions that are as maintainable as they are novel.
11 years of coding experience
23 years of employment as a software developer
MS Applied Mathematics, MS Applied Mathematics at École Polytechnique
The University of Sydney
MS Operations Research, MS Operations Research at ENSTA
Mathematics & Physics, Mathematics & Physics at Lycée Sainte-Geneviève
English, French