← Back Experience & Education

Experience

06/2025 – 02/2027
Working Student, Short-Term Trading
E.ON  ·  Munich
  • Developed ML models for wind and solar power forecasting (gradient boosting, neural networks, regularized regression, Optuna) combining multiple numerical weather prediction sources; reduced RMSE by 20 % vs. operational forecasts.
  • Built automated Azure data pipelines for forecasts and actuals at 15-minute resolution, with quality checks and alerting.
  • Developed and operated dashboards (Plotly Dash, Streamlit) used daily by traders to monitor and compare forecasts; containerized with Docker and deployed via Azure DevOps CI/CD on Kubernetes.
11/2022 – 02/2025
Working Student, Innovation Lab New Materials
Infineon Technologies AG  ·  Regensburg
  • Built an ML model classifying materials by their FTIR spectra for reverse engineering.
  • Co-designed experimental setups for novel semiconductor materials; analyzed and presented the results.

Education

02/2027 – 07/2027
(planned)
Semester Abroad
Universidade do Porto  ·  Porto
Exchange semester to conclude my M.Sc. Physics
04/2025 – 07/2027
(expected)
M.Sc. Physics — Current Grade 1.5 (German scale, 1.0 = best)
Ludwig-Maximilians-Universität  ·  Munich
AI in Physics certificate

Thesis  Extrapolation of Nonlinear Dynamics in Multidimensional Parameter and Physical Space with Next Generation Reservoir Computing (paper in preparation)

Institution  German Aerospace Center (DLR)  ·  04/2026 – 01/2027

Supervisor  Dr. Christoph Räth

Focus  Data-driven forecasting of chaotic and spatiotemporal systems with next generation reservoir computing (NGRC). For chaotic systems (Lorenz, Rössler, Chua, Thomas), NGRC extrapolates across a two-dimensional parameter plane into dynamical regimes unseen in training, validated via Lyapunov exponents and correlation dimension. For spatiotemporal systems (2D Kuramoto–Sivashinsky, 2D complex Ginzburg–Landau), it generalizes without retraining to larger domains and, for Ginzburg–Landau, to unseen parameter regimes.

Electives  Machine Learning (1.3) · Computational Physics (1.0) · AI in Particle Physics (1.3) · Complex Systems and AI (1.0) · Bayesian Inference and Machine Learning (1.3) · AI Lab: time series forecasting with quantum reservoir computing (1.3)

10/2021 – 02/2025
B.Sc. Physics — Grade 1.1
Universität Regensburg  ·  Regensburg
Minor: Chemistry

Thesis  Preconditioning for Wilson–clover fermions using gauge-equivariant deep neural networks

Institution  High Energy Physics Group, Universität Regensburg

Supervisor  Prof. Dr. Tilo Wettig

Focus  ML acceleration of QCD solvers: gauge-equivariant neural networks (CNNs) as preconditioners for computationally intensive solvers.

Electives  Quantum Statistics · Advanced Quantum Mechanics

09/2015 – 07/2021
Abitur — Grade 1.1
Veit-Höser-Gymnasium  ·  Bogen
DPG Prize (physics) and DMV Prize (mathematics) for the top graduate in these subjects