I take on a small amount of client work.

Everything published on this site is the portfolio: the benchmarks are run by me, the pricing data is collected by my own pipelines, and the recommendations are the same ones I would give a client. If that way of working fits a problem you have, I am open to hearing about it.

The work I am good at, and have done for German enterprises in health insurance, banking, medical associations, and product data: turning messy real-world documents into structured, validated data; choosing between cloud APIs, self-hosted models, and simpler systems based on measurements rather than preference; and building the pipeline once the evidence says which route is worth building. Background: senior AI engineer, PhD in physics (RWTH Aachen / Forschungszentrum Jülich), mostly open-source tooling, comfortable deploying on your infrastructure when compliance requires it.

There is no fixed catalog and no rate card. Describe the decision or the system, and you will hear early and honestly whether I am the right person for it. If the honest answer is that you do not need me, or do not need AI at all, that is the answer you will get.