Germany’s Prior Labs raises €9M to bring the AI revolution to tabular data

Freiburg-based Prior Labs, a German AI startup developing foundation models for spreadsheets and databases, has raised €9M in a pre-seed round of funding. 

The round was led by Balderton Capital, with participation from XTX Ventures, Hector Foundation, Atlantic Labs and Galion.exe.

Prominent AI investors like Thomas Wolf (Founder & CSO, Hugging Face), Peter Sarlin (Founder & CEO, Silo AI), Guy Podjarny (Founder, Snyk & Tessl), Ed Grefenstette (Director, DeepMind), Robin Rombach (Founder & CEO, Black Forest Labs), Chris Lynch (Founding Investor Data Robot & CEO, AtScale), Ash Kulkarni (CEO, Elastic) and other business leaders also participated.

The funds will support Prior Labs’ product development, team expansion, and broader adoption of its technology.

James Wise, Partner at Balderton Capital, says, “Tabular data is the backbone of science and business, yet the AI revolution transforming text, images and video has had only a marginal impact on tabular data – until now.”

“Prior Labs’ breakthrough gives everyone the super-powers of machine learning without needing to train their own models on their own data. We’re thrilled to support this world-class team as they redefine how industries unlock the value of their data.”

What does Prior Labs offer?

Tabular data—structured data in tables, spreadsheets, and databases—is essential across industries like healthcare, finance, and manufacturing, yet AI advancements in this field have lagged behind those in text and images. Prior Labs’ TabPFN model provides a universal approach to tabular data analysis.

Trained on 130 million synthetic datasets, it identifies patterns in any dataset without task-specific training. As a foundation model, it supports fine-tuning with proprietary data to improve accuracy and adaptability for real-world applications.

Frank Hutter, co-founder and CEO of Prior Labs, says, “Most of the world’s critical decisions are powered by tabular data, yet tools to analyse it are outdated and lacking. We’re bringing a quantum leap to the predictions businesses can make from their most valuable data and building a future where engaging with tables is as seamless as using AI for text or images. We can deliver faster, more accurate predictions that empower businesses to do more with less.”

A recent Nature paper reports that TabPFN outperformed “state-of-the-art models” in over 96 per cent of cases on small tabular data. It achieves the same accuracy as the next best model with 50 per cent less data and delivers better performance in 2.8 seconds compared to existing models that take over four hours. It can be applied to any dataset with minimal code.

TabPFN improves decision-making in trading, finance, and business analytics by providing faster and accurate predictions. In data-limited fields like healthcare, medicine, and climate science, the same results are achieved with 50 per cent less data, enabling scientific research and discovery.

What’s next for Prior Labs?

Prior Labs was founded in late 2024 by Professor Dr. Frank Hutter, Noah Hollmann, and Sauraj Gambhir. The team, with over 20 years of combined machine learning experience, developed a foundation model for tabular data called TabPFN. Their work, published in Nature, highlights the model’s potential to transform data analysis.

Prior Labs is now scaling its impact by integrating its API into business data workflows, helping businesses unlock the full potential of their tabular data.

The company is improving its model’s speed, accuracy, and efficiency, adding support for text features, fine-tuning with proprietary data, and incorporating contextual information to enhance predictions.