Product Engineer
Pumas-AI Inc., Italy
My name is Lorenzo Contento, PhD and I have been working in product development at PumasAI since 2023. I explore new ways to integrate machine learning and AI with traditional pharmacometrics modelling and implement them in the DeepPumas software suite. I have employed neural networks to model systems whose dynamics are partly unknown, using random effects to individualize the learned laws to each subject and to link different observables. Recently I have started evaluating normalizing flows as a way to describe complex random effect distributions and their relationship to covariates.
Before moving to industry, I spent several years doing academic research in applied mathematics, with a focus on mathematical modelling. I received my Bachelor’s and Master’s degrees in mathematics from the University of Udine, respectively in 2010 and 2012. Under the guidance of Prof. Masayasu Mimura, in 2016 I received my PhD in mathematics from Meiji University, Tokyo, where I stayed on for 4 additional years as a postdoctoral researcher. There I studied pattern formation in ecological systems modelled by partial differential equations, mainly relying on numerical simulation and bifurcation methods. Later I worked at the University of Bonn in the group of Prof. Jan Hasenauer where I got acquainted with the field of computational biology. My research focused on uncertainty quantification in mathematical models for the spread of epidemics in the context of the ORCHESTRA Covid-19 Response project under the Horizon 2020 EU funding program.