Summary
Paul Albert is an applied scientist and postdoctoral-trained researcher with eight years of experience developing efficient, low-supervision solutions for vision and language models. Now at Amazon after a fellowship at the Australian Institute for Machine Learning, he focuses on parameter-efficient fine-tuning, knowledge composition, continual learning and recommender systems that scale to global customers. His background spans academic research (PhD work on unsupervised and noisy-label vision), teaching algorithms and ML, and industry-facing projects that bridge drone/agricultural imaging to large-scale GenAI applications. Notably, he combines rigorous publication-driven research with practical system-building—optimizing models under label scarcity and noisy data to make them production-ready.
8 years of coding experience
2 years of employment as a software developer
Scientific Baccalauréat (S), STEM, High honours, Scientific Baccalauréat (S), STEM, High honours at Lycée Pierre Paul Riquet - St-Orens de Gameville, France
Preparation in Math, Physics and Computer Science., Preparation in Math, Physics and Computer Science. at Lycée Déodat de Séverac - Toulouse, France
Masters degree, Artificial Intelligence, Masters degree, Artificial Intelligence at CentraleSupélec - Paris, France
Master's degree, Human-computer interactions, machine learning, semantic web, Master's degree, Human-computer interactions, machine learning, semantic web at Lorraine University - Nancy, France
French, Spanish, English