Niccolò Dal Santo

I am a deep learning software engineer at MathWorks, where I enhance MATLAB Deep Learning Toolbox with new AI techniques. My main responsibilities include deep learning for physics simulations (PINNs, Neural ODEs), generative models (such as GANs), RNNs (GRUs) and CNNs deep learning architectures. I am also engaged in a scientific research collaboration with University of Massachusetts Amherst, where we build in MATLAB efficient deep learning turbulence emulators for engine simulations and integrate them into the OpenFOAM C++ code.

Previously, I was Postdoctoral researcher in the group of Prof. S. Deparis at EPFL, in Lausanne, Switzerland, where I worked on developing novel deep learning methods to solve parameter-dependent PDEs. In 2018 I obtained my PhD in Computational Mathematics at the same institution under the supervision of Prof. A. Quarteroni and Dr A. Manzoni, with a thesis on preconditioning methods for finite element simulations of parametrized partial differential equations.

Interests: Machine learning, Deep learning, Computational physics, Model order reduction, Numerical optimization, High-performance computing.

Education:

PhD in Computational Mathematics, EPFL, 2018

MS in Mathematical Engineering, Politecnico di Milano, 2014

BS in Mathematical Engineering, Politecnico di Milano, 2012