01 — About
About me
I'm a machine-learning researcher working mostly on brain signals. I try to keep my analyses honest, especially when the data is small.
I'm a computational researcher based in Dresden. As a research assistant at TU Dresden, I build graph neural networks that help classify ADHD from EEG. My master's thesis, done at DZNE, looked at stress and resilience in brain recordings from high-density microelectrode arrays.
Before research, I worked in industry for a few years — fraud detection, credit risk, and fintech — which taught me a lot about making models that work on real data.
In my spare time I run a small Linux server at home and automate small things around it.
Research interests
what drives meNeural decoding & dynamics
Using graph neural networks and transformers to make sense of high-dimensional brain recordings.
Multimodal integration
How the brain combines inputs from different senses over time, using time-series and state models.
Computational stability
How neural circuits keep working reliably in the presence of noise — drawing on anomaly detection and control theory.
Interpretability in AI
Understanding what deep models actually learn, and connecting them back to biology.
Education
degreesSkills
toolboxPython · expertMATLAB · expertSQLC / C++JavaRGAMSlinear algebragraph theoryoptimization
PyTorchPyTorch GeometricTensorFlowTransformersViT / MAEGNNunsupervised clusteringXAI · SHAP / LIME
MNE-Pythontime-frequency analysisspectral parameterizationwaveletsphase-locking valuetime-series analysis
LinuxDockerGitAzure MLHPC / Slurm
English · nativeHindi · nativeGerman · A2Arabic · basic
Recognition
notedContact
reach meHappy to talk about research, collaborations, or positions.