Summary
Mathieu Orhan is an AI Research Engineer based in Paris with nine years’ experience building production-ready ML and computer vision systems. He has progressed from deep learning scientist roles into senior research and manager positions, leading core research teams and delivering high-impact engineering such as 10x inference speedups and large-scale training and inference infrastructure on AWS and TPUs. Mathieu combines hands-on modelling—residual error analysis, synthetic data engines, multi-view geometric reasoning and GNNs—with ML Ops expertise to turn research into robust, explainable products that enabled full automation and multi-million-dollar savings. He has a strong track record in document understanding and OCR (boosting text detection from ~60% to 99% on noisy Japanese faxes) and in multi-view vehicle damage assessment. Trained at École normale supérieure Paris-Saclay and École des Ponts, he blends rigorous academic foundations with pragmatic product-driven experimentation. Colleagues would note his appetite for principled problem framing and for improving tooling and pipelines that raise team velocity and reproducibility.
9 years of coding experience
6 years of employment as a software developer
Master 2 (M2), Master MVA, High honors, Master 2 (M2), Master MVA, High honors at École normale supérieure Paris-Saclay
Mathematics and Computer Science, Machine Learning and Computer Vision, Mathematics and Computer Science, Machine Learning and Computer Vision at École des Ponts ParisTech
Mathématiques et physique, Mathématiques et physique at CPGE Lycée Bellevue
English, French