I design and build data platforms for large-scale, real-time, and ephemeral workloads - from distributed object stores and disaggregated storage to high-throughput processing pipelines in hybrid cloud environments.
I’ve built open-source systems for shuffle storage and ephemeral data handling, and contributed to performance evaluations of emerging runtime architectures. My work covers data product strategy, ML infrastructure, and taking systems from proof-of-concept to production.
Before that, I led engineering on interactive visualization tools, developed weather modeling software for HPC clusters, and ported cryptographic libraries to embedded devices. A common thread across my work has been making systems faster and more reliable.
I hold an MSc in Computer Science from ETH Zürich and like to explore new systems design patterns through independent research.
A selection of systems and tools I’ve built is on the projects page.
This is my personal, occasionally updated website. Views are my own.
Find me on GitHub or reach out at me @ thisdomain.