Building Foundation Models for Tabular Data with Prior Labs

Sauraj Gambhir, Co-founder of Prior Labs, joined us at SLUSH 2025 to discuss how they're bringing the transformer revolution to tabular data. Born from research challenges in medical machine learning, Prior Labs has developed TabPFN — a foundation model that can handle any tabular dataset without needing to build custom models for each use case. With nearly 2 million downloads and applications ranging from railway maintenance to lung disease diagnostics, their open-source approach and focus on community is transforming how organizations work with structured data across healthcare, finance, energy and beyond.