Today, we are thrilled to announce that Oak HC/FT has led the $225M Series A for Fundamental, an AI company that has built a powerful Large Tabular Model to drive predictions from enterprise data.
More than 80% of enterprise data lives in tables, yet LLMs are built for processing data like language, images, and code. LLMs, as a result, are excellent at language-based reasoning and workflow automation. However, they are poorly suited for the multidimensional, non-sequential nature of tabular data – the exact data that serves as the "backbone of enterprise decision-making.”
Fundamental is solving the gap left by other AI labs with the introduction of NEXUS, its Large Tabular Model purpose-built for structured data. Trained on billions of tabular datasets, NEXUS is able to transform how companies extract insights from their data. By creating a transformer architecture specifically for numerical and relational data, Fundamental allows enterprises to move beyond analyzing the past to answering critical forward-looking questions. And, unlike legacy methods like gradient boosting (XGBoost), which require extensive manual feature engineering and lack "transfer learning" capabilities, NEXUS automatically learns the underlying patterns and dependencies of raw data.
The model’s ability to deliver PhD-level insights with minimal preprocessing means companies can achieve faster time-to-value across a range of use cases – from predicting financial fraud and hospital readmissions to forecasting demand and energy prices.
To accelerate this vision, Fundamental has entered into an unprecedented strategic partnership with Amazon Web Services (AWS). Fundamental is now a first-party "seller of record" on the AWS dashboard, allowing any AWS customer to deploy NEXUS with a single line of code, bypassing typical multi-month procurement cycles. This collaboration validates Fundamental’s technology as a core pillar of the future AI infrastructure stack.
Our conviction in Fundamental starts with its CEO and co-founder, Jeremy Fraenkel, a leader who possesses a rare combination of technical depth and commercial execution. Jeremy earned a graduate degree in Machine Learning from UC Berkeley, providing him with the rigorous technical grounding required to lead a frontier AI research lab. However, Jeremy is more than just a technologist; he brings a level of enterprise and financial literacy that is uncommon in the world of AI research. Having built a strong finance background through roles at JPMorgan and Bridgewater, Jeremy understands the mission-critical nature of the data Fundamental aims to transform. The research team is led by former DeepMind researchers that have been at the forefront of deep learning and LTM research.
Our conviction is further strengthened by what we see as an inflection point in enterprise AI adoption, particularly across sectors we know well, financial services and healthcare. Recent advances in LLMs have unlocked significant value and efficiency in specific workflows and manual operations. However, in data-heavy industries such as healthcare and financial services, LLMs alone have struggled to realize the full value embedded in enterprise data, particularly when it comes to generating predictions and outcomes that result in hard ROIs. We believe that LTMs have now reached a critical threshold, across research maturity, technical capability, data availability, and enterprise readiness, that makes it possible to unlock this value by combining the world’s data context with enterprise-specific datasets. Fundamental is bringing this capability to life at the enterprise level.
While Oak HC/FT is traditionally known for investing in vertical-specific solutions within fintech and healthcare, our lead investment in Fundamental represents a new focus on highly strategic horizontal bets in foundational AI infrastructure to supplement the vertical focus. Although Fundamental is building a horizontal "prediction engine" designed to function as a third pillar of AI alongside text and vision, we are excited about the immediate and high impact applications in the two large verticals of financial services and healthcare, which sit squarely within our areas of deepest expertise. Jeremy and his team have created a technology that can revolutionize fraud detection, credit underwriting, or macro-aware price forecasting in financial services, as well as patient risk stratification and claims analytics in healthcare.
We are excited to leverage Oak’s deep industry networks and regulatory fluency to help Jeremy unlock these complex GTM motions, helping Fundamental move beyond the pilot phase to become the default predictive operating system for the world’s most critical industries. We are honored to support Jeremy Fraenkel and the Fundamental team as they engineer the future of enterprise decision-making.