Applied Mathematician · Scientific Machine Learning Researcher
I develop numerical methods for PDEs, fluid dynamics, and inverse problems, with a focus on reduced-order modelling and scientific machine learning. My work combines mathematical analysis, reproducible simulation, and experience building machine-learning and data systems in industry.
Website · CV · LinkedIn · Email
Numerical simulation · Inverse problems and uncertainty · Reduced-order modelling · Scientific machine learning
A Fourier pseudo-spectral Navier–Stokes solver for periodic flows, with dealiasing, energy and enstrophy diagnostics, and analytical verification. Manufactured-solution studies demonstrate fourth-order time accuracy for the full-order solver.
A Python package for reconstructing missing greyscale image regions while preserving known pixels exactly. Includes a command-line interface and comparisons with harmonic and biharmonic reconstruction across five images and nineteen masks. Best suited to simple structures and thin missing regions; fine textures and large holes remain challenging.
Knowledge distillation for financial-sentiment classification, with saved baseline and student evaluations. The smaller model achieves 96.9% test accuracy with approximately 19% fewer parameters, compared with 97.6% baseline accuracy on the recorded Financial PhraseBank split.
An application developed at Intuos Srl for exploring recorded aircraft and flight data. Connects FastAPI and IBM DB2 to a React/TypeScript interface with interactive maps, authentication, rule-based flight-envelope checks, and Docker Compose configuration. Designed for internal analysis with authorised database access.
Explore more projects and research →
- M.Sc. Mathematical Engineering, University of L’Aquila, Italy, 2021.
- B.Sc. Mathematics, Obafemi Awolowo University, Nigeria, 2018.
Thesis and research background →
Tools: Python, NumPy, SciPy, PyTorch, scikit-learn, MATLAB, FEniCSx, PETSc, SQL, FastAPI, Docker, Git.
Based in L’Aquila, Italy. I welcome conversations about scientific computing, applied ML, and uncertainty-aware models for physical systems.