Backed by $255M, Fundamental emerges with AI model built for business data

What's the deal? San Francisco AI startup Fundamental emerged from stealth on February 5, 2026, announcing $255 million in total funding—a $30 million seed round and a $225 million Series A—at a unicorn valuation. Founded by DeepMind alumni, it launched what it claims is the first publicly available Large Tabular Model (LTM), designed to predict from the structured, spreadsheet-like data underpinning business.

Why now? The launch capitalises on a significant gap in the AI market. While massive investment has flowed into large language models (LLMs) for text and images, the trillions of dollars in business value locked in enterprise tables remained largely untouched by deep learning advancements. Traditional machine learning methods for this data require slow, manual work from data scientists, creating a bottleneck for companies wanting predictive insights.

What could go wrong? As a new and highly specialised category of AI, the long-term performance and adaptability of Large Tabular Models against evolving enterprise data challenges remain unproven. Furthermore, the model’s deterministic design, while ideal for regulated industries, means it cannot process unstructured data like text, limiting its use where a generalised AI is needed.

The signal: Fundamental’s launch and substantial backing signal a major new frontier in applied AI, moving beyond generative tools to predictive, decision-making engines. The pre-secured Fortune 100 contracts and strategic AWS partnership demonstrate intense enterprise demand to modernise core data analysis and indicate the next wave of AI value will come from automating complex business forecasting.

Source:
Financial Times
TechCrunch
FinSMEs
The Tech Buzz
Business Wire

A.M.