Zurich, Zurich, Switzerland
Full spectrum data specialist with deep grasp of the underlying math & technology. From the top of the stack (ML/AI, visualisation, dashboards, apps, ...) to the bottom (hardware, OS, DBs, containers, clusters, cloud infra, ...), with competence in a broad range of languages, libraries & tools in between, I strive to enable companies to extract more value from their data & build scalable solutions for the future. Quick to adapt to new concepts, technologies and programming languages depending on the needs of the project.
Selected projects: - Technical lead for an application running on a large-scale cluster with ~10k internal users at a leading global bank, coordinating clean-up of several years of tech-debt, while delivering new features on time with a team of 5+ people. - Drive a customer analytics platform from conception to realisation (multi-year project at a large retail bank, role: conception, elaboration, architecture, development end-to-end), featuring spark ETL generator framework incl. bitemporal data tracking & containerised interface for data scientists with automated promotion from laboratory to factory, incl. fully-featured UI, unified handling of R & python - Help a transportation provider to analyse and reduce maintenance costs (machine learning and convex optimization for cost driver analysis) - Developing a (Python) matching library/API for a bank’s data science team - Machine-learning projects in healthcare and transport sectors - Information architecture analysis for an energy provider - Overhaul of automated (Java) testing framework for DWH of large insurance
Caring about python packaging in general, library patterns, testing best practices and balancing the long tail of support vs. (volunteer) maintenance effort and CI health.
Maintainer of 100+ feedstocks; numpy & scipy, BLAS, LLVM, SDL stacks, OpenSSL, pyarrow, protobuf, abseil tensorflow, etc. Recipient of Google's "Open Source Peer Bonus" (2023)
Aside from conda-forge & scipy, contributions to various projects trackers, many upstreams from respective conda-forge feedstocks (e.g. cython, cvxpy, faiss, allennlp), ~100 pandas PRs, loosely participating in container ecosystem.