Postdoctoral researcher at the Institute of Materials Science, TU Darmstadt. I work on AI for materials science, connecting first-principles simulation with semantic data: density functional theory and Bayesian optimization on one side, EMMO-based ontologies and knowledge graphs on the other.
Research
- Materials ontologies. Domain ontologies within EMMO that give electronic, atomistic, mesoscopic and continuum simulations of magnetic materials a common description.
- Active learning and Bayesian optimization. Automated construction of effective Hamiltonians, and multi-objective optimization over high-dimensional alloy composition spaces.
- High-throughput first-principles workflows. EMTO-CPA and VASP pipelines for high-entropy alloys and functional materials.
Selected projects
| Project | What it is |
|---|---|
| commagmo | Computational Magnetic Materials Ontology: an EMMO-aligned OWL 2 ontology for multiscale magnetic simulations |
| HEA | High-throughput EMTO-CPA database, ML surrogates and multi-objective BO for soft-magnetic high-entropy alloys |
| BO | Active-learning construction of the effective Hamiltonian of BaTiO₃ |
| mdo-repro | Reproduction of the Materials Design Ontology, with competency-question tests on Materials Project data |
| texwatch | Local LaTeX preview: incremental build on save, browser refresh that keeps your scroll position |
Open to research positions in AI for materials science from February 2027.