About Me
I am a postdoctoral researcher at École Polytechnique Fédérale de Lausanne at the Chair of Biostatistics held by Mats Stensrud.
I am broadly interested in developing statistical methods to advance data analysis under realistic assumptions, particularly in causal inference.
Before joining EPFL, I completed my PhD degree at the Statistical Laboratory of the University of Cambridge under the supervision of Qingyuan Zhao and obtained a Master in Mathematics at TU Munich advised by Mathias Drton.
Beyond this, I also undertook industry projects in the implementation of RNNs with Tensorflow, Bayesian optimal experimental design and detection of synthetically lethal gene pairs based on CRISPR knock-out experiments.
- Causality
- Sensitivity Analysis
- Randomization Inference
- Selective Inference
- Applied Statistics
PhD Mathematics of Information, 2024
University of Cambridge
M.Sc. Mathematics, 2020
Technical University of Munich
Exchange Research Student, 2020
Kyoto University
Visiting Student, 2018
University of Oxford
B.Sc. Mathematics, 2017
Ludwig Maximilian University, Munich
34th Conference of the Austro-Swiss Region of the Biometric Society (ROeS)
Graz, Austria
Selective Randomization Inference for Adaptive Experiments
International Indian Statistical Association Conference
Lincoln, NE, USA
Optimization-based Sensitivity Analysis
Advanced Methods and Data Science Seminar
Novartis, online
Selective Randomization Inference for Adaptive Experiments
Causal Machine Learning Workshop
University of Southampton
Selective Randomization Inference for Adaptive Experiments
Machine Learning and Data Science Seminar
Okinawa Institute of Science and Technology, Japan
Sensitivity Analysis with the R2-Calculus
European Causal Inference Meeting
Oxford, UK
Randomization Inference with Concentration Inequalities
IMS International Conference on Statistics and Data Science
Sevilla, Spain
Randomization Inference with Concentration Inequalities
International Conference on Robust Statistics
Stresa, Italy
Selective Randomization Inference for Adaptive Experiments
Bernoulli-IMS 11th World Congress in Probability and Statistics
Bochum, Germany
Selective Randomization Inference for Adaptive Experiments
American Causal Inference Conference
Seattle, WA, USA
Selective Randomization Inference for Adaptive Studies
European Causal Inference Meeting
Copenhagen, Denmark
Selective Randomization Inference for Adaptive Experiments
Response-Adaptive Randomisation in Clinical Trials Workshop
MRC Biostatistics Unit, Cambridge
Selective Randomization Inference for Adaptive Clinical Studies
Statistics Student Seminar
University of Chicago
Selective Randomization Inference for Adaptive Studies
During my time in Cambridge, I was twice teaching assistant for the Part III course ‘Causal Inference’, mainly holding classroom-style tutorials. At EPFL, I had the same role for the Master course ‘Computation and Visualization’, marked the students’ programming assignments and answered their questions in the weekly tutorials. Moreover, I acted as “supervisor” for multiple Bachelor statistics courses at Cambridge. In this role, I supervised several small groups of students, marked their homework assignments and tailored the biweekly tutorials to their individual progress.
I have also gathered experience advising students on individual projects. During my PhD, I supervised one summer exchange student visiting the University of Cambridge, and at EPFL I supervised two Bachelor theses this far.
In summer 2026, and I were project supervisors in the ‘Young Researchers in Mathematics Program’ held at EPFL’s Bernoulli Center. A small group of undergraduate students were invited to come to Lausanne for a week and work on their first research project. Nils and I advised the students in the statistics track of the programme and we keep working on the project with the students.
In the autumn semester 2026, I teach a PhD course called ‘Perspectives on Randomness in Statistics’ at EPFL. We investigate different frameworks of conceptualizing randomness (frequentist super-population, design-based, Bayesian etc.) and discuss their suitability for specific applications.