Fundamental, a San Francisco, CA-based AI company building a Large Tabular Model to drive predictions from enterprise data, has raised $255 million in a Series A funding round led by Oak HC/FT.
The round also saw participation from Valour Equity Partners, Battery Ventures, Salesforce Ventures, and Hetz Ventures, as well as angel investors, including Perplexity co-founder and CEO Aravind Srinivas, Wiz co-founder and CEO Assaf Rappaport, Brex co-founder Henrique Dubugras, and Datadog co-founder and CEO Olivier Pomel.
This funding round follows a $30M Seed round.
Read More:Goodfire Raises $150M in Series B Funding
The company plans to use the funds to increase computing capacity, expand enterprise deployments, and quickly grow its team in research, engineering, and go-to-market.
Fundamental comes out of stealth with $255M in funding and introduces NEXUS, a Large Tabular Model designed to transform enterprise data into predictive insights and forecasts.
Fundamental, an AI company specializing in Large Tabular Models (LTMs) for enterprise data, has partnered with Amazon Web Services (AWS) to speed adoption of its model among AWS customers. Starting today, AWS customers can purchase and deploy NEXUS directly in their AWS environment, including compute and storage, using AWS’s secure, scalable infrastructure. At the same time, Fundamental has secured seven-figure contracts with Fortune 100 companies using the model for predictive tasks such as demand forecasting, price prediction, and customer churn.
“Fundamental’s structured data prediction model builds on AWS’s advanced AI tools, helping enterprise customers analyze tabular data at scale,” said Dave Brown, VP of Compute, Platforms & ML Services at AWS. “By partnering with Fundamental, we make it easy for customers to turn tabular data—the core of enterprise decision-making—into powerful predictive insights. This collaboration shows our commitment to delivering transformative AI solutions with the security and scalability enterprises need.”
Enterprises have long relied on outdated machine learning algorithms—developed before deep learning—when analyzing data and making predictions. While recent advances in deep learning have focused on LLMs and other models designed for unstructured data like text, images and video, these models aren’t well suited for tabular data. They struggle with the non-sequential, nonlinear relationships in tables and often can’t handle enterprise-scale datasets because of their size and complexity. As a result, they fail to extract value from the tabular data that drives key business decisions.
Built by DeepMind alums, Fundamental’s first publicly available Large Tabular Model (LTM), NEXUS, lets enterprises make predictions with unprecedented accuracy. NEXUS replaces outdated predictive analytics with a foundation model designed specifically for tabular data. It helps companies go beyond analyzing past events to answer forward-looking questions—such as what will happen next, when risks may arise, or where opportunities lie—while delivering fast results and enterprise-ready deployment across any cloud.
Built from billions of tabular datasets and trained on Amazon SageMaker HyperPod, NEXUS captures the complex, nonlinear relationships across rows and columns. Enterprises can easily integrate it into their existing data stacks—often with just one line of code. Once connected, NEXUS ingests raw tabular data and automatically learns its structure patterns, and dependencies without manual feature engineering or training. The result is a model that delivers far more accurate predictions than traditional machine learning methods.
“The importance of Fundamental’s model can’t be overstated—structured, relational data hasn’t yet fully benefited from the deep learning revolution,” said Annie Lamont, Co-Founder & Managing Partner at Oak HC/FT. “Fundamental can predict everything from financial fraud to hospital readmissions to energy prices, making it valuable across nearly every industry. With a world class research team that combines deep technical expertise with strong commercial execution, the company offers a rare blend of scientific rigour and enterprise know how. We’re proud to partner with them on this journey.”
About Fundamental
Founded in 2024, and led by CEO Jeremy Fraenkel, Fundamental is shaping the future of enterprise decision-making. Co-founded by DeepMind alums, the company developed NEXUS—its most powerful Large Tabular Model (LTM)—designed specifically for structured data that drives trillions of dollars in business value. While most AI companies focus on text and images, Fundamental focuses on the tabular data that powers real enterprise decisions. Backed by top investors and trusted by Fortune 500 companies, Fundamental unlocks trillions in value by giving businesses the power to predict.
Read More:LearnWell Receives Investment from Goldman Sachs Alternatives
