Building ML for neuroscience and machines that act in the real world.

Dual-study B.Sc. student in Data Science & AI at DHBW Ravensburg. I work on graph learning for connectomes, autonomous racing, and sleep neurotechnology.

Currently

Selected work

All projects →

Two-panel figure. Panel a: an annotated reference trains a GNN on the unlabeled target's anonymous wiring; the GNN's rare-type candidates are graph-matched to reference rare-type neurons. Panel b: dot plot of accuracy per source-target pair, showing two-stage typing above the GNN alone on all five pairs.

Research

Flybrain: typing rare neurons in an unlabeled connectome

Rare-type accuracy 22 → 41% on a held-out brain (BANC), in a pre-registered test.

ML
Neurotech

In progress

GFR Driverless: autonomous stack for Formula Student

TODO: one-line outcome, e.g. a completed event or a lap-time / cone-detection result.

Autonomous Driving
Robotics

In progress

EEG sleep band

Goal: stimulation timed to slow oscillations, evaluated offline before the first hardware prototype.

Neurotech
ML

Notes

All notes →

3 Nov 2026
Devlog #0: Why I'm building an EEG sleep band
Placeholder post that shows the note format: why this project, what comes first, and how progress will be logged.

Contact

I am open to research collaborations, Master's opportunities, and conversations with people building deep-tech companies.

GitHub: Paradoxsmile
LinkedIn: vogtmaximilian
ORCID: 0009-0005-4171-3538