Francesco Romor
Francesco Romor

Postdoctoral researcher

About Me

I am a computational scientist with hands-on experience in high-performance computing, scientific machine learning and numerical simulations for real-world engineering problems. I develop and deploy scalable numerical solvers (C++, MPI, GPU via Kokkos/CUDA) and AI/machine learning/data-driven models (surrogate models, generative models). My work targets concrete applications, most recently patient-specific aortic blood flows and data assimilation from medical images.

Download Academic CV Download CV
Interests
  • Machine learning
  • Numerical modelling
  • High-performance computing
Education
  • PhD Mathematical analysis, modelling and applications

    International School for Advanced Studies (SISSA)

  • MS Mathematics

    University of Triest

  • BSc Mathematics

    University of Triest

Preprints
Publications
(2026). Shape-informed graph neural networks and data assimilation: application to velocity and pressure reconstruction in aortic blood flow. SIAM journal on Imaging Sciences.
(2025). Efficient numerical strategies for entropy-regularized semi-discrete optimal transport. Computer Methods in Applied Mechanics and Engineering.
(2025). Explicable hyper-reduced order models with autoencoders. Journal of Computational Physics.
(2023). Generative models for the deformation of industrial shapes with linear geometric constraints: Model order and parameter space reductions. Computer Methods in Applied Mechanics and Engineering.