Geophysicist with a focus on machine learning @ GFZ Potsdam;
Working on understanding faults across scales
Sometimes also working on (Bio)NLP, paleontology, excitation spectra or gene regulatory networks
Research Interests
I'm a researcher focussing on deciphering the dynamics of faults zones on time-scales from earthquakes to long-term tectonic deformation.
With my combined background in mathematics, computer science and seismology, I employ and develop methods to derive precise insights from large-scale geophysical data.
In particular, I focus on machine learning and big data methods with a strong passion for open science and open source.
Through my research I want to extend our knowledge of Solid Earth processes and ultimately help disaster preparedness and mitigation.
General Assembly of the European Seismological Commission, Istanbul, TurkeyKeynote
Ohio State University, Columbus, USA
2025
EPOS-Seismology workshop, Athens, Greece
Institut Terre et Environnement de Strasbourg, Strasbourg, France
Kyoto University, Kyoto, Japan
NIED, Tsukuba, Japan
NUTS-AI workshop on artificial intelligence and application to solid Earth, Paris, France
JpGU meeting, session "Science of slow-to-fast earthquakes", virtual contribution
2024
AGU Fall Meeting, session "Advances in Theoretical and Computational Seismology", Washington DC, USA
ERC TECTONIC seminar
UC Santa Cruz, Santa Cruz, USA
Stanford University, Stanford, USA
Berkeley Seismological Laboratory, Berkeley, USA
Lamont-Doherty Earth Observatory, Palisades, USA
GFZ German Research Centre for Geosciences, Potsdam, Germany
Institut de Physique du Globe de Paris, Paris, France
2023
National Academies Committee on Solid Earth Geophysics, online
University of Washington, Seattle, USA
Massachusetts Institute of Technology, Cambridge, USA
Universidad de Chile, Santiago de Chile, Chile
2022
Third European Conference on Earthquake Engineering and Seismology (3ECEES), Bucharest, Romania
8th International SCADS.AI Summer School, Leipzig, Germany
Freie Universität Berlin, Berlin, Germany
Karlsruhe Institute of Technology, Karlsruhe, Germany
2021
Hosted full-day workshop on SeisBench at Istituto Nazionale di Geofisica e Vulcanologia, online and Pisa, Italy
Kick-off focus group "AI for Natural Disaster management" of the International Telecommunication Union
University of Oxford, Oxford, UK (online)
Université Grenoble Alpes, Grenoble, France (online)
Air Force Technical Applications Center (online)
Open Source
SeisBench is an open source toolbox for machine learning in seismology. It provides
the building blocks for developing and applying machine learning models to seismic waveform
data, including a large repository of pre-trained models, and is used in numerous
earthquake monitoring workflows. I'm one of the initial developers of SeisBench and the lead
maintainer.
A high-throughput seismic phase associator. It compiles earthquake catalogs from raw seismic
phase picks. It is designed to handle the large pick volumes
produced by modern deep learning based pickers and dense networks.
I developed PyOcto to create high-resolution earthquake catalogs with millions of picks.