Luigi Carratino
Sr Research Engineer at Microsoft Research – AI for Science
Based in Berlin
I am a Senior Research Engineer on the AI for Science team at Microsoft Research, working on Material Discovery.
I completed my Ph.D. in 2020 at the Laboratory for Computational and Statistical Learning, University of Genova, under the supervision of Lorenzo Rosasco, where my research focused on developing scalable kernel methods and efficient Gaussian process optimization algorithms. During my doctoral studies, I interned at Google Brain (now Google DeepMind), where I had the opportunity to collaborate with Moustapha Cissé, Rodolphe Jenatton, and Jean-Philippe Vert. There, I established the theoretical foundations of the Mixup data-augmentation technique for deep learning.
Following my Ph.D., I spent a year and a half as a Postdoctoral Researcher at the Machine Learning Genoa Center, developing computationally efficient ML algorithms blending randomized linear algebra, statistics, and GPU optimization. I then joined Zalando as an Applied Scientist (2021–2022), where I designed and deployed causal discovery and inference solutions to measure and optimize marketing impact. I then joined yahoo! dsp research as a Senior Research Scientist (2022–2026), where I led ML initiatives in ad ranking, prediction modeling, bid optimization, and supply quality assessment for programmatic advertising.
My research interests span large-scale kernel methods, deep learning, multi-armed bandits, and Bayesian optimization.