Prior Labs
Research Scientist, Foundational Data Science
Berlin +2 · On-site · Full-time · Posted Jul 2, 2026
- Location
- Berlin
Freiburg
New York
About the job
Build foundational tools, datasets, and benchmarks for tabular foundation models, guiding research toward strong performance on real-world data-science problems.
What you'll do
- Develop frontier tools and methods that extend TabPFN, including thinking, scaling, and agentic capabilities.
- Set research priorities, incorporate external research and customer needs, and publish results.
- Create trustworthy benchmarks, implement baseline and competitor models, and build an automated human-in-the-loop pipeline for scaling data and benchmark development.
What they ask for
- Demonstrated ability to solve data-science problems across diverse domains and datasets, optimizing across suites of tasks.
- Broad ML toolbox, including strong results with gradient-boosted trees such as XGBoost, not only deep learning.
- Understanding of dataset defects such as leakage, label noise, distribution shift, duplication, and mislabeled targets, and how they affect training or benchmarks.
- Ability to thrive as a senior individual contributor in an ambiguous, early-stage, low-process environment, with sound judgment on complex data-science work.
Nice to have
- Experience building or extending evaluation harnesses, benchmark suites, or experiment frameworks.
- Experience building LLM- or agent-assisted human-in-the-loop pipelines.
Tools
TabPFN, XGBoost, LLM
AI focus
Research and develop tabular foundation models, including their thinking, scaling, and agentic capabilities; build agentic tools and data pipelines.
Model-written summary · the original posting is authoritative