AI startup Fundamental has emerged from stealth mode after raising a whopping $255 million in total funding, including a $225 million Series A round.
The round was led by Oak HC/FT with participation from Valor Equity Partners, Battery Ventures, Salesforce Ventures, and prominent angels like Perplexity CEO Aravind Srinivas, valuing the company at an impressive $1.2 billion.
Fundamental's Nexus: Pioneering Structured Data AI
At the core of Fundamental's innovation is Nexus, a Large Tabular Model (LTM) designed specifically for analyzing massive structured datasets like enterprise tables with billions of rows.
Unlike traditional large language models from OpenAI or Anthropic that excel in unstructured data but falter on large spreadsheets due to context window limits, Nexus uses a novel non-transformer architecture for deterministic results.
This breakthrough addresses longstanding challenges in big data analysis, evolving from early tools like Hadoop in the 2000s to today's AI-driven approaches that still require armies of data scientists.
Strategic Partnerships and Early Traction
Fundamental has already inked seven-figure contracts with Fortune 100 clients, demonstrating immediate enterprise demand for its technology.
A key partnership with AWS allows users to deploy Nexus directly from existing instances, accelerating adoption across cloud environments.
CEO Jeremy Fraenkel emphasized, “With our model Nexus, we have built the best foundation model to handle structured data,” enabling one model for diverse use cases with superior performance.
Looking ahead, Nexus promises to democratize big data insights, reducing reliance on specialized teams and unlocking new efficiencies for industries like finance and healthcare.
As AI investments surge in 2026, Fundamental's funding underscores investor confidence in specialized models tackling unstructured data gaps.
The company's rapid rise from stealth to unicorn status signals a transformative shift in how enterprises harness their vast structured data reserves.
Maria Lourdes
