Summary
Arthur Herbout is a data scientist with a decade of experience building production recommender systems and applied ML across retail, media, healthcare, and research labs. Currently an Associate in Data Science & Advanced Analytics at Cerberus after leading personalization efforts at Bluecore, he combines rigorous academic training from Columbia and CentraleSupélec with hands-on engineering to lift model performance in production. His work spans collaborative filtering, cold-start solutions, computer vision, and neural architecture search, with wins such as a 12% uplift on iHeartRadio’s podcast recommender. Comfortable bridging research and product, he has taught recommender systems at Columbia and optimized models for hardware-constrained environments early in his career. Fluent in translating math-heavy ideas into deployable systems, he often surfaces practical gains from principled methods. Based in New York, he brings a rare mix of elite French math training and broad industry ML product experience.
10 years of coding experience
5 years of employment as a software developer
CentraleSupélec
Scientific baccalauréat, mathematics, Obtained with High Honors, Scientific baccalauréat, mathematics, Obtained with High Honors at Lycée La Rochefoucauld, Paris, France
Master of science, Data science, Master of science, Data science at Columbia University
Intensive preparation in Mathematics and Physics for the highly competitive national entrance exams, Mathematics, physics, chemistry, Intensive preparation in Mathematics and Physics for the highly competitive national entrance exams, Mathematics, physics, chemistry at Lycée Henri IV, Paris, France
French, Spanish, English