Fundamental Lands $255M Series A for Innovative Big Data Analysis

AI laboratory Fundamental officially launched from stealth on Thursday, introducing a novel foundation model designed to address the long-standing challenge of extracting valuable insights from the vast volumes of structured data generated by businesses. The company asserts that by integrating established predictive AI methodologies with modern tools, it can revolutionize the way large enterprises approach data analysis.

"While Large Language Models (LLMs) have demonstrated remarkable efficacy in processing unstructured data such as text, audio, video, and code, their performance with structured data like tables is significantly limited," CEO Jeremy Fraenkel explained to TechCrunch. He added, "Through our model, Nexus, we have engineered the premier foundation model specifically tailored to manage this particular data type."

This innovative approach has already garnered substantial investor attention, with Fundamental emerging from stealth with an impressive $255 million in total funding. The majority of this capital stems from a recent $225 million Series A round, spearheaded by prominent firms including Oak HC/FT, Valor Equity Partners, Battery Ventures, and Salesforce Ventures, with additional participation from Hetz Ventures. Angel investments further contributed, notably from Perplexity CEO Aravind Srinivas, Brex co-founder Henrique Dubugras, and Datadog CEO Olivier Pomel.

Designated as a Large Tabular Model (LTM) instead of a Large Language Model (LLM), Fundamental's Nexus diverges from current AI methodologies in several key respects. Notably, the model operates deterministically, ensuring consistent outputs for identical queries. Furthermore, it eschews the transformer architecture that underpins models from most modern AI laboratories. Fundamental classifies Nexus as a foundation model due to its adherence to standard pre-training and fine-tuning stages, yet the ultimate outcome offers a distinctly different proposition compared to solutions from providers like OpenAI or Anthropic.

These distinctions are crucial as Fundamental targets a specific use-case where existing AI models frequently encounter limitations. Transformer-based AI models, constrained by their context window, often struggle with reasoning across exceptionally vast datasets, such as spreadsheets containing billions of rows. Given that such immense structured datasets are prevalent within large enterprises, this presents a substantial market opportunity for models capable of operating at this scale.

Fraenkel perceives this as an immense opportunity for Fundamental. He believes that Nexus empowers the company to introduce advanced, contemporary techniques to Big Data analysis, providing solutions that are both more potent and adaptable than the algorithms presently employed.

"With Nexus, you can now deploy a single model across all your operational use cases, enabling a massive expansion in the scope of problems you can address," he conveyed to TechCrunch. He further emphasized, "For each of these use cases, you will achieve superior performance compared to what could be accomplished even with a large team of data scientists."

This compelling promise has already translated into numerous high-profile engagements, including multi-million dollar contracts with Fortune 100 clients. Additionally, Fundamental has forged a strategic partnership with AWS, which will enable AWS users to seamlessly deploy Nexus directly from their existing cloud instances.