Scientific computing for complex real-world questions.

About

I am pursuing a PhD in Physics at the University of Vienna while working as a Scientist at the Center for Modelling and Simulation, University for Continuing Education Krems. I study permanent magnets and build computational methods and software as part of the Christian Doppler Laboratory for magnet design through physics-informed machine learning.

My background spans Biology and Computational Science. It informs a cross-disciplinary approach to complex natural-science questions, bringing together data analysis, modeling, simulation, and software development.

Interests

I am interested in how computation can help address the materials challenges behind more sustainable energy technologies. The prospect of improving permanent magnets while reducing their reliance on rare earths is especially compelling to me.

I care about the trustworthiness of machine learning as much as its predictive power. Questions about when a model is reliable enough to guide an optimization decision sparked my interest in uncertainty quantification. From there, I dived into using uncertainty-aware models to make optimization more efficient, as well as active-learning approaches, which I first explored in my master’s thesis. I believe that uncertainty estimation is a key ingredient for making machine learning more efficient.

I also enjoy developing software tools that make my life and that of my colleagues easiers. During my Master's project I started developing a Python package that has become the backbone of my optimization workflows ever since. I believe computational tools for science need to be reliable and easy to use. Qualities such as modularity, reproducibility, and maintainability are the foundation of good scientific software, and I strive to implement them in my work.