I am a computational science engineer with a PhD in mathematics, experienced in scientific software development, numerical simulation, AI integration and leadership.
Plexim · Zurich · 2024 - present
I lead the scientific software team at Plexim, which works on core solver technologies and strategic R&D for next-generation simulation tools for power-electronic systems design and validation. My own focus at the moment is the native AI integration of PLECS, our power-electronics simulator, so that agents can analyze, create and simulate multi-physics models directly in the editor.
ETH Zurich · 2026 - present
Teaching Numerical Methods for Partial Differential Equations, a core course of the CSE program and part of the first generation of mandatory ETH courses graded by AI. I also co-supervise graduate students.
Team Lead Research Scientist · 2023 - 2024
Scientific Software Developer · 2022 - 2023
I joined Plexim in 2022 as a research scientist and software engineer, working on numerical methods and modeling for circuit simulation in C++. I was quickly promoted to team leader and put in charge of developing PLECS Spice, a circuit solver for semiconductor device-level simulation, written from scratch for highly non-linear, mixed-signal, multiphysics circuits.
The project was part of the PowerizeD consortium, funded by Horizon Europe (€1M). The new solver shipped at the end of 2025. For 2026, I wrote a follow-up proposal to develop AI tools that shorten the design cycle of power-electronic systems, which was awarded a second €1M by the same institution (Moore4Power).
Alongside the solver, I developed numerical schemes for real-time simulation on the RT Box, our real-time simulator, and a new parameter identification method for thermal networks. I also give technical workshops at semiconductor manufacturers internationally and bring what they need back into the product.
As coding agents became part of our daily work, I built the CLI Plexim developers use to run them in isolated microVMs. Today my main project is the native AI integration of PLECS.
PhD student · 2018 - 2022
My research was on numerical methods for electromagnetism, using finite and boundary element methods. My thesis developed boundary integral equations for Hodge operators (Laplace, Dirac) on manifolds, leading to discoveries that were warmly received at the conference Fast Boundary Element Methods in Industrial Applications and earned a nomination for the ETH Medal in 2022. My collaboration with Prof. Ralf Hiptmair continues to this day.
I also co-led the development of NumPDE, a C++ repository of 100+ simulation algorithms for fluid flow, electromagnetism, conservation laws, heat propagation and more. I still maintain it, and it remains the standard framework for the Numerical Methods for PDEs course at ETH, which I now teach.
R&D Intern · 2017
I built a Levenberg-Marquardt nonlinear least-squares optimizer with geodesic acceleration. It was applied to identify surrogate models of heat propagation in large-scale HVAC systems and to parametrize robotic arms.
M.Sc. & B.A. · 2012 - 2018
I hold an M.Sc. in Applied Mathematics (2018) and a B.A. in Honours Mathematics and Computer Science (2016) from McGill University. I graduated with first class honours and appeared on the Dean's Honour List (top 10% of students in my faculty). In my Master's thesis, I proved the first convergence result for Discrete Exterior Calculus (DEC) and implemented a DEC solver. Despite its popular use in computer graphics and computational topology, understanding the method's convergence in general remains an open problem.
French and English speaker. Always open to discussing new challenges in computational science.