I build simulation software where Python ergonomics meets Rust performance — with physics underneath.
I am a Research Software Engineer with a PhD in Computational Physics, working at the intersection of numerical simulation, high-performance scientific software, and machine learning.
My work usually starts with a physical problem and ends somewhere between a numerical method, a simulation framework, an optimized kernel, and a Python API that other researchers can actually enjoy using.
physical model
│
▼
numerical formulation
│
▼
simulation software
Python ⇄ Rust ⇄ HPC
│
▼
analysis / ML / discovery
- 🧮 Scientific computing — finite elements, boundary elements, finite differences, ODE integration and multiphysics simulation
- 🦀 Python + Rust — PyO3 bindings,
ndarray,rayon,cubecl,maturin, parallel CPU workloads and GPU-oriented development - 🤖 Machine learning for simulation — surrogate models, autoencoders, dimensionality reduction, clustering and Bayesian hyperparameter optimisation
- 🧲 Computational magnetism — micromagnetics, magnetisation dynamics, spin waves and three-dimensional magnetic systems
- 🛠️ Research software engineering — testing, CI/CD, packaging, reproducible workflows and open-source maintenance
A growing collection of tools for scientific computing and simulation.
A Python library for defining numerical approximations of scalar and vector fields, developed with finite-element analysis in mind.
A finite-element micromagnetic simulation framework with a Python domain-specific interface and computationally intensive components implemented in Rust.
The design philosophy is simple:
Keep the scientific interface expressive. Put the heavy lifting where it belongs.
|
|
|
|
| Area | Examples |
|---|---|
| Micromagnetics | 3D nanomagnets, non-centrosymmetric ferromagnets, bulk DMI, magnetisation dynamics, spin-wave dynamics |
| Scientific ML | High-dimensional simulation data, convolutional/variational autoencoders, clustering and physics-informed surrogate models |
| Multiphysics FEM | Current-density + heat diffusion simulations in superconducting devices for optimisation of current pad position and geometry |
| Simulation frameworks | Open-source Python simulation ecosystems, numerical field representations and Python/Rust hybrid architectures |
| High-performance computing | Moving computational kernels from Python to Rust and exploiting CPU/GPU parallelism |
I enjoy the part of research software engineering where physics, mathematics and software design have to agree with each other.
I also supervise and mentor researchers working on:
- micromagnetic simulations,
- unsupervised learning for high-dimensional simulation data,
- numerical simulation software in Python and Rust,
- CPU and GPU acceleration of scientific workloads.
PhD in Computational Physics
University of Strasbourg, France
Integrated Master of Science in Physics
Indian Institute of Technology Roorkee, India
My professional work has taken me through research environments in Germany, the United Kingdom, France and India, with a strong focus on interdisciplinary and open-source scientific software.
scientific computing · Rust for science · Python APIs · FEM · BEM · micromagnetics · simulation software · machine learning for physics · HPC · open source
If any of those overlap with what you're working on, feel free to reach out.