Fundamental, an AI company that has built its most powerful Large Tabular Model (“LTM”) to drive predictions from enterprise data, announced it is emerging from stealth with $255M in funding. Founded in October 2024, the company raised a $30M Seed round and a $225M Series A round led by Oak HC/FT, with participation from top tier venture capital firms including Valor Equity Partners, Battery Ventures, Salesforce Ventures, and Hetz Ventures. Notable angel investors include Perplexity co-founder and CEO Aravind Srinivas, co-founder and CEO of Wiz Assaf Rappaport, Brex co-founder Henrique Dubugras, and Datadog co-founder and CEO Olivier Pomel. The new capital will be used to scale compute, expand enterprise deployments, and rapidly grow the team across research, engineering, and go-to-market.
“We’ve built a generalized foundation model specifically to leverage the world’s most valuable data: the billions of tables that underpin predictions in every enterprise, across every vertical,” says Fundamental CEO & co-founder Jeremy Fraenkel. “NEXUS is the OS for business decisions.”
Fundamental has also entered into a strategic partnership with Amazon Web Services (AWS) to accelerate enterprise adoption of its model to AWS customers. Starting today, AWS customers can buy and deploy NEXUS in their AWS environment in the same way they buy compute or storage on their AWS dashboard – all backed by AWS’s secure, reliable, and scalable infrastructure. In parallel, Fundamental has secured seven-figure contracts with Fortune 100 enterprises that are applying this model to predictive use cases including demand forecasting, price prediction, and customer churn.
Also Read: VPN vs Zero Trust Security: Why Enterprises Are Rethinking Remote Access in 2026
“Fundamental’s structured data prediction model builds on AWS’s advanced AI offerings, helping enterprise customers fill a crucial gap in comprehensive tabular data analysis at scale,” said Dave Brown, VP of Compute, Platforms & ML Services at AWS. “By partnering with Fundamental, we are making it seamless for customers to transform tabular data – the backbone of enterprise decision-making – into a powerful predictive asset. This collaboration exemplifies our commitment to bringing transformative AI solutions to market with the enterprise-grade security and scalability our customers demand.”
Enterprises have historically relied on antiquated machine learning algorithms that predate deep learning to analyze their data, inform decisions, and make predictions. In contrast, recent advances in deep learning have largely centered on LLMs and related architectures optimized for unstructured, sequential data such as text, images, and video. As a result, these models are poorly suited to capture the non-sequential, nonlinear relationships inherent in tabular data, and struggle to process enterprise-scale tables at all due to size and dimensionality constraints. This means they’re not built to derive value from the tabular datasets that inform every critical enterprise decision.
Built by DeepMind alumni, Fundamental’s first publicly-available LTM, NEXUS, gives enterprises the power to predict with far greater accuracy than ever before. NEXUS replaces legacy predictive analytics with a purpose-built foundation model designed specifically for tabular data. Fundamental enables enterprises to move beyond analysis of past events to answer forward-looking questions like what will happen next, when risks will emerge, or where opportunities exist -all with fast time-to-value and enterprise-grade deployment on any cloud infrastructure.
Built from the ground-up on billions of tabular datasets and trained on Amazon SageMaker HyperPod, NEXUS understands non-linear relationships and interactions that exist across rows and columns. Enterprises can integrate NEXUS directly into their existing data stacks with minimal effort, often with a single line of code. Once connected, the model ingests raw tabular data and automatically learns the underlying structure, patterns, and dependencies without extensive feature engineering or manual training. The result is an engine that delivers significantly more accurate predictions than traditional machine learning methods.
“The significance of Fundamental’s model is hard to overstate – structured, relational data has yet to see the benefits of the deep learning revolution,” said Annie Lamont, Co-Founder & Managing Partner at Oak HC/FT. “Fundamental’s ability to predict anything from financial fraud to hospital readmission to energy prices positions the company to support virtually every industry and sector. With a world-class research team that blends deep technical expertise with proven commercial execution, the company brings a rare mix of research rigor and enterprise GTM understanding. We’re honored to be partnering with them on their journey.”
SOURCE: Buisnesswire