senior ml engineer @ eghed, a machine learning consultancy in gothenburg.
i build large-scale ml systems that make it from research into production — most recently self-supervised vision transformers (dinov2) trained on 250 tb of railway imagery for trafikverket.

phd in distributed deep learning from kth, 2025 — thesis. before eghed, ml research scientist at rise (2019–2025) and visiting researcher at nyu with kyunghyun cho, with papers at tmlr, icml workshops, and ieee big data.

federated & decentralized learning · self-supervised learning · ood generalization · representation learning · ml for climate & sustainability

a few things i’ve built: publications · industry projects · a göteborgsvarvet results explorer from 40k finish times.

get in touch — email or linkedin. supervising a thesis? see students.