Fundamental raises $255M Series A with a new take on big data analysis

An artificial intelligence lab called Fundamental came out of stealth on Thursday, introducing a new type of foundation model designed to address a long-standing challenge for enterprises: extracting meaningful insights from massive volumes of structured data. By blending traditional predictive AI techniques with modern approaches, the company says it can change how large organisations analyse and use their data.

“While LLMs have been great at working with unstructured data, like text, audio, video, and code, they don’t work well with structured data like tables,” Jeremy Fraenkel, the company’s chief executive, told TechCrunch. “With our model Nexus, we have built the best foundation model to handle that type of data.”

The concept has already attracted substantial investor interest. Fundamental is emerging from stealth with $255 million in total funding and a $1.2 billion valuation. Most of that capital comes from a $225 million Series A round led by Oak HC/FT, Valour Equity Partners, Battery Ventures, and Salesforce Ventures. Hetz Ventures also participated in the round, alongside angel investors including Aravind Srinivas, Henrique Dubugras, and Olivier Pomel. Fundamentally, Nexus is a large tabular model (LTM), rather than a large language model (LLM), marking a clear departure from prevailing AI design philosophies. Nexus is deterministic, meaning it produces the same answer every time it is asked the same question, and it does not rely on the transformer architecture that underpins most modern AI models. While Fundamental still describes Nexus as a foundation model — because it undergoes standard pre-training and fine-tuning processes — the result is fundamentally different from what enterprises typically receive when working with labs such as OpenAI or Anthropic.

These distinctions matter because Fundamental is targeting a problem area where many contemporary AI systems struggle. Transformer-based models are limited by their context windows, which restrict how much data they can process at once. As a result, they often fail to reason effectively with large live structured datasets, such as spreadsheets with billions of rows. Yet this type of data is commonplace inside large enterprises, creating a significant gap in existing AI capabilities.

Fraenkel believes this gap represents a significant opportunity for Fundamental. With Nexus, the company aims to apply modern AI techniques to large-scale data analysis, delivering tools that are more adaptable and powerful than the algorithms enterprises typically rely on today.

“You can now have one model across all of your use cases, so you can now expand massively the number of use cases that you tackle,” Fraenkel told TechCrunch. “And on each one of those use cases, you get better performance than what you would otherwise be able to do with an army of data scientists.”

That value proposition has already translated into commercial traction. Fundamental says it has secured several high-profile deals, including seven-figure contracts with Fortune 100 companies. The startup has also formed a strategic partnership with Amazon Web Services, enabling AWS customers to deploy the Nexus model directly from their existing cloud environments.