The announcement landed on a Wednesday morning, timed perhaps for maximum effect: Fundamental, a San Francisco startup barely a year old, had raised $225 million in Series A funding. The round valued the company at somewhere between $1.2 billion and $1.4 billion—sources differ, as they often do in these situations—catapulting it into unicorn territory faster than most founders dare to dream.
What makes the February 5, 2026 deal particularly striking isn't just the velocity or the valuation. It's the thesis. While much of Silicon Valley remains fixated on chatbots and image generators, Fundamental is building what CEO Jeremy Fraenkel calls a "Large Tabular Model"—essentially, a foundation model for the spreadsheets and databases where enterprises actually store their most valuable information.
"Enterprises' most valuable data lives in tables," Fraenkel told reporters, articulating a problem that sounds almost mundane until you consider its scope. "And LLMs are terrible at structured data tasks."
Oak HC/FT led the Series A, joined by Salesforce Ventures, Valor Equity Partners, Battery Ventures, and Hetz Ventures. The investor roster reads like a who's-who of enterprise tech: Perplexity's Aravind Srinivas, Brex's Henrique Dubugras, Datadog's Olivier Pomel, and Wiz's Assaf Rappaport all participated as angels. Combined with a $30 million seed round raised shortly after the company's October 2024 founding, Fundamental has now amassed $255 million in total capital.
For context, Crunchbase data shows the median U.S. Series A round in 2024-2025 hovering somewhere between $11 million and $20 million. Fundamental raised roughly twenty times that figure.
The Amazon Angle
The funding came wrapped in a strategic partnership with Amazon Web Services—the kind of announcement that suggests careful choreography. Starting immediately, AWS customers can purchase Fundamental's NEXUS platform directly through their AWS dashboard, with Amazon acting as seller-of-record via its Marketplace. The model runs on Amazon SageMaker HyperPod and can be deployed entirely within customer environments, a detail that matters considerably for enterprises wary of sending sensitive data outside their walls.
"Fundamental's innovative approach to tabular data analysis at scale aligns with our mission to help customers unlock the full value of their data," said Dave Brown, AWS vice president of Compute, Platforms & ML Services, in the company's press release. (Translation: AWS sees enough potential here to grant prime distribution real estate.)
VentureBeat confirmed the AWS partnership details, though neither company disclosed financial terms.
A Different Kind of Foundation Model
NEXUS isn't built on transformers—the architecture powering ChatGPT, Claude, and virtually every other generative AI system making headlines lately. Fundamental claims its approach is deterministic, trained on what it describes as "billions of tabular datasets" to handle enterprise prediction tasks: demand forecasting, price prediction, churn analysis, fraud detection, hospital readmission risk, equipment failure.
The pitch centers on simplicity. One-line integration, the company says, though veterans of enterprise software deployments might raise an eyebrow at that promise.
Fraenkel previously co-founded Arkifi, an AI finance workflow startup that raised $9 million from Khosla Ventures and Nyca in 2023. His chief science officer, Marta Garnelo, comes from DeepMind. Founding advisor Wojciech Czarnecki also spent time at DeepMind and Isomorphic Labs. Israeli tech press reports a third co-founder, Gabriel Suissa, with R&D operations in Israel, though the company's website doesn't publicly list team members—a curious omission for a unicorn courting enterprise buyers.
Early Revenue, Big Questions
Fundamental claims it has already locked in "multiple seven-figure contracts" with Fortune 100 companies across oil and gas, finance, and healthcare, according to the Financial Times. The company employs approximately 35 people, per VentureBeat—a lean team for a $1.2 billion valuation, even in an era of AI-inflated multiples.
The capital will go toward scaling compute infrastructure, purchasing training data, and expanding headcount. Standard startup playbook stuff, though the sheer size of the round suggests something more ambitious. Perhaps Fundamental is trying to move fast enough to establish a category before competitors—or the hyperscalers themselves—decide tabular data deserves dedicated foundation models.
The company positions itself against both traditional machine learning toolchains like XGBoost and LightGBM, and the broader data infrastructure stack dominated by Databricks, Snowflake, and the cloud providers' own platforms. That's a crowded field with entrenched players.
What It Means
Whether "Large Tabular Models" become a recognized category or fade into the noise of AI's endless subcategories remains an open question. Salesforce Ventures, one of Fundamental's backers, doubled its AI investment fund to $1 billion in 2024—a sign of how much capital is chasing this space, and how difficult it is to predict which bets will pay off.
For now, Fundamental has secured two critical advantages: Fortune 100 customers willing to pay seven figures, and AWS distribution that puts its product in front of cloud buyers already accustomed to clicking "purchase" for infrastructure services. Whether that's enough to justify a unicorn valuation built on less than 18 months of operating history is a question that will take years to answer.
But in Silicon Valley, where patience has never been a virtue and foundation models have become the new gold rush, Fundamental has at least succeeded in making structured data look sexy. That alone might be worth $225 million to the right investors.
