James Wise 
- Portfolio News
- 02 June, 2026
Following news of SAP’s intent to acquire Prior Labs, James Wise reflects on the journey to date.
We first met Frank, Noah and Sauraj before they formally founded the business through our shared interest in tabular foundation models. After speaking to people working in the field across the globe, we knew instantly that Prior Labs was the team to back.
As we wrote when we led their seed (and only) investment round last year, Frank Hutter has already changed the field of data science once with his work on AutoML. With Noah and Sauraj now onboard, we felt they had the technical and operational capability to build the best model in the world. Within a month of our investment a paper on TabPFN, their series of tabular models, was published in Nature becoming the most referenced paper on AI last year. What surprised even us though was the rapid pace at which the team kept improving their models, with the release of TabPFN 2.0, and then scaling mode, putting them in a league of their own.
They achieved this both by building one of the most talent dense teams in AI globally, but also by taking a pragmatic approach to working with the machine learning community. They started with an opensource approach before launching their API and then commercial partnerships, improving the speed and accuracy of predictions in fields as diverse as cancer diagnosis to rail analytics in the process.
We are extremely excited about Prior Labs’ future. Most of the world’s valuable data still sits in spreadsheets and databases, and being able to replace manually-tuned, gradient-boosted trees with a one-shot model approach will transform machine learning for non-experts, in the same way LLMs have for unstructrued data.
It’s been a pleasure working with them over the last 18 months. We’ve been fortunate to support them in hiring a truly exceptional global team, and I am sure it’s just the beginning of what they can achieve.
Watch my interview with Professor Hutter below:
