Fundamental, an AI lab focused on extracting insights from massive volumes of structured enterprise data, has raised $255M and emerged from stealth with a new foundation model designed specifically for large-scale tabular analysis.
Founded by Jeremy Fraenkel, Fundamental is tackling a long-standing limitation in enterprise analytics. While large language models excel at unstructured data such as text, audio, and code, they struggle with structured data like tables and relational datasets. Fundamental’s answer is Nexus, a foundation model purpose-built to work with enterprise-scale tabular data.
Nexus is designed to reason over datasets that extend far beyond the context limits of transformer-based models, enabling analysis of data volumes that are common inside large organizations but difficult for modern AI systems to handle.
Large Tabular Models instead of large language models
Rather than relying on transformer architectures, Fundamental has developed what it calls a Large Tabular Model. Nexus is deterministic, producing consistent results for the same query, and follows a pre-training and fine-tuning process similar to foundation models while delivering fundamentally different capabilities.
This approach allows enterprises to analyze extremely large structured datasets, including tables with billions of rows, without fragmenting data or relying on complex pipelines of specialized tools.
Advertisement
Backed by major investors at unicorn valuation
The $255M total funding includes a $225M Series A round valuing the company at $1.2B. The round was led by Oak HC/FT, Valor Equity Partners, Battery Ventures, and Salesforce Ventures, with participation from Hetz Ventures. Angel investors include Aravind Srinivas, Henrique Dubugras, and Olivier Pomel.
The scale of the round reflects investor confidence in Fundamental’s vision to modernize big data analytics using a single, general-purpose model rather than fragmented predictive systems.
One model across multiple enterprise use cases
Fundamental positions Nexus as a unifying layer for enterprise analytics. Instead of deploying separate models and workflows for forecasting, anomaly detection, optimization, and reporting, organizations can apply one model across a wide range of use cases.
This consolidation is intended to reduce reliance on large data science teams while improving performance across analytics tasks that traditionally require extensive manual tuning.
Early traction with Fortune 100 customers
Fundamental reports early commercial momentum, including seven-figure contracts with Fortune 100 enterprises. The company has also entered a strategic partnership with AWS, enabling customers to deploy Nexus directly within their existing AWS environments.
This integration allows enterprises to apply the model to live datasets without re-architecting their infrastructure.
About Fundamental
Fundamental is an AI lab building foundation models for structured enterprise data. Its Large Tabular Model, Nexus, is designed to analyze massive datasets deterministically and at scale, helping large organizations extract insights from complex data environments without relying on fragmented tools or transformer-based limitations.
