AI research and benchmarking
Technical tasks, rubrics, and benchmark frameworks for frontier models, together with failure-mode analysis, scientific coding, and review at PhD/postdoctoral level.
Physics | AI Research | Technical Project Leadership
I work where experimental physics, scientific coding, and applied AI meet. I am most interested in tasks where technical depth and a useful result have to come together.
Germany | German and English | Remote and on-site by arrangement
I am an experimental physicist by training and enjoy working where scientific questions, technical systems, and practical decisions meet. My path has moved from detector development, data analysis, and teaching into applied AI research and model evaluation.
In current AI projects I help plan and improve tasks, workflows, rubrics, and benchmark frameworks for frontier models. This includes guardrails for evaluation integrity, ethical and technical safety, confidentiality, and reliable tool use.
Three areas connect my work so far. In practice, the boundaries between them are usually quite fluid.
Technical tasks, rubrics, and benchmark frameworks for frontier models, together with failure-mode analysis, scientific coding, and review at PhD/postdoctoral level.
Structuring complex work, connecting technical and economic constraints, and turning an open question into a workflow that produces reliable results.
Experimental astroparticle physics, detector and sensor systems, data analysis, teaching, and collaboration across research, engineering, and industry.
A few direct links. The protected CV provides the full context.
For recruiters and project partners
The full CV, including professional history, publications, teaching, project context, and PDF versions, is intentionally not indexed publicly. An agreed password opens both the English and German versions.
Already received access? Go straight to the login. A short email is enough to request access.