Vincenzo Di Florio

I’m Vincenzo Di Florio, a physicist turned computational scientist — driven by a simple ambition: turning the language of physics into algorithms that actually work on real, messy problems.

I hold a Ph.D. in Pure and Applied Mathematics from Politecnico di Torino, with a background that blends analytical rigor and numerical modeling. I’m currently a Post-doctoral Research Fellow at MOX – Modeling and Scientific Computing, Politecnico di Milano.

My main research focus is biomolecular electrostatics — in particular the Poisson–Boltzmann equation, which describes how charged molecules behave in electrolytic environments. This work led to NextGenPB, a high-resolution finite element solver with analytical super-resolution that I developed from the ground up.

Alongside that, I’m actively working on Physics-Informed Neural Networks (PINNs) and Physics-Guided Neural Networks (PGNNs) applied to electrostatics — approaches that embed physical laws directly into machine learning models, enabling accurate predictions with limited data and enforcing physical consistency by design.

I also work on non-equilibrium statistical mechanics, studying how complex systems evolve far from equilibrium — with the ambition of connecting rigorous theoretical results to realistic models.

If you are interested in collaborations in computational physics, scientific machine learning, or high-performance scientific software, feel free to reach out. You can browse my publications and find my code on GitHub.