The Spreadsheet Revolution: How AI Finally Cracked Business’s Most Valuable Data

In the breathless race to build ever-more-sophisticated AI systems, a curious blind spot has persisted: the humble spreadsheet. While Large Language Models have mastered human conversation and image generators have conquered digital art, the structured, relational data that forms the backbone of global commerce—the rows and columns of ERP systems, CRMs, and financial ledgers—has remained stubbornly resistant to the deep learning revolution.

Until now.

Today, Fundamental, a San Francisco-based AI firm founded by DeepMind alumni, emerges from stealth mode with a staggering $255 million in total funding to announce NEXUS, a Large Tabular Model (LTM) that promises to transform how enterprises predict business outcomes.

The Problem With Numbers

“The most valuable data in the world lives in tables,” declares Jeremy Fraenkel, CEO and Co-founder of Fundamental, “and until now there has been no good foundation model built specifically to understand it.”

This isn’t merely an academic observation. Consider the typical enterprise scenario: A retail company wants to predict which customers are at risk of churning. A bank needs to identify fraudulent transactions in real-time. A hospital must forecast which patients require urgent follow-up care. These aren’t abstract problems—they’re the daily challenges that determine whether businesses thrive or fail.

Yet current AI approaches struggle mightily with tabular data. The reason? Numbers.

When Large Language Models process text, they tokenize words into manageable chunks. But when faced with numbers, they apply the same logic, breaking “2.3” into three separate tokens: “2,” “.”, and “3.” This seemingly minor quirk has profound consequences. As Fraenkel explains, “That essentially means you lose the understanding of the distribution of numbers. It’s not like a calculator; you don’t always get the right answer because the model doesn’t understand the concept of numbers natively.”

Beyond Sequential Logic

The challenge runs deeper than tokenization. Enterprise data is inherently non-sequential. A customer’s likelihood to churn isn’t simply a timeline—it’s a complex intersection of transaction frequency, support ticket sentiment, and regional economic conditions.

Traditional machine learning approaches have required armies of data scientists to manually define features—the specific variables a model should consider. This bespoke, labor-intensive process can take months for each use case. NEXUS aims to eliminate this entirely.

Trained on billions of real-world tabular datasets using Amazon SageMaker HyperPod, NEXUS is designed to ingest raw tables directly. It identifies latent patterns across columns and rows that human analysts might miss, effectively reading the hidden language of the grid to understand non-linear interactions.

The Order-Invariance Problem

Here’s where things get particularly interesting. Language is inherently order-dependent—the sequence of words matters immensely. But tabular data is order-invariant. As Fraenkel illustrates with a healthcare example: “If I give you a table with hundreds of thousands of patients and ask you to predict which of them has diabetes, it shouldn’t matter if the first column is height and the second is weight, or vice versa.”

While LLMs are highly sensitive to word order, NEXUS is architected to understand that shifting column positions should not impact the underlying prediction. This fundamental architectural difference enables the model to truly understand the structure of business data.

Not Just Another Spreadsheet Tool

Recent high-profile integrations, such as Anthropic’s Claude appearing directly within Microsoft Excel, have suggested that LLMs are already solving the spreadsheet problem. But Fraenkel draws a crucial distinction: “What they are doing is essentially at the formula layer—formulas are text, they are like code. We aren’t trying to allow you to build a financial model in Excel. We are helping you make a forecast.”

NEXUS operates at the predictive layer, designed for split-second decisions where humans aren’t in the loop. While tools like Claude can summarize a spreadsheet, NEXUS is built to predict the next row—whether that’s an equipment failure in a factory or the probability of a patient being readmitted to a hospital.

The Implementation Revolution

Fundamental’s approach to market adoption is as innovative as its technology. The company has already secured several seven-figure contracts with Fortune 100 organizations, facilitated by a strategic go-to-market architecture where Amazon Web Services serves as the seller of record on the AWS Marketplace.

This allows enterprise leaders to procure and deploy NEXUS using existing AWS credits, treating predictive intelligence as a standard utility alongside compute and storage. For engineers, the experience is high-impact but low-friction: NEXUS operates via a Python-based interface at a purely predictive layer rather than a conversational one.

Developers connect raw tables directly to the model and label specific target columns—such as credit default probability or maintenance risk score—to trigger the forecast. The model then returns regressions or classifications directly into the enterprise data stack, functioning as a silent, high-speed engine for automated decision-making.

Beyond the Bottom Line

While the commercial implications of demand forecasting and price prediction are clear, Fundamental emphasizes the societal benefits of predictive intelligence. The company highlights several areas where NEXUS can prevent catastrophic outcomes:

Infrastructure Safety: By analyzing sensor data and maintenance records, NEXUS can predict failures like pipe corrosion. The company points to the Flint water crisis—which cost over $1 billion in repairs—as an example where predictive monitoring could have prevented life-threatening contamination.

Healthcare Optimization: During the COVID-19 crisis, PPE shortages cost hospitals $323 billion in a single year. Fundamental argues that by using manufacturing and epidemiological data, NEXUS can predict shortages 4-6 weeks before peak demand, triggering emergency manufacturing in time to save lives.

Climate Resilience: NEXUS aims to provide 30-60 day flood and drought predictions, such as for the 2022 Pakistan floods which caused $30 billion in damages.

Patient Outcomes: The model is being used to predict hospital readmission risks by analyzing patient demographics and social determinants. As the company puts it: “A single mother working two jobs shouldn’t end up back in the ER because we failed to predict she’d need follow-up care.”

Performance vs. Latency

In the enterprise world, the definition of “better” varies by industry. For some, it’s speed; for others, it’s raw accuracy. Fraenkel explains: “In terms of latency, it depends on the use case. If you are a researcher trying to understand what drugs to administer to a patient in Africa, latency doesn’t matter as much. You are trying to make a more accurate decision that can end up saving the most lives possible.”

For financial institutions, even marginal improvements translate to massive value. “Increasing the prediction accuracy by half a percent is worth billions of dollars for a bank,” Fraenkel notes. “For different use cases, the magnitude of the percentage increase changes, but we can get you to a better performance than what you have currently.”

The $255 Million Vote of Confidence

The $225 million Series A, led by Oak HC/FT with participation from Salesforce Ventures, Valor Equity Partners, and Battery Ventures, signals high-conviction belief that tabular data is the next great frontier. Notable angel investors including leaders from Perplexity, Wiz, Brex, and Datadog further validate the company’s pedigree.

Annie Lamont, Co-Founder and Managing Partner at Oak HC/FT, articulates the sentiment: “The significance of Fundamental’s model is hard to overstate—structured, relational data has yet to see the benefits of the deep learning revolution.”

The AWS Partnership Advantage

Fundamental’s strategic partnership with Amazon Web Services provides crucial infrastructure advantages. NEXUS is integrated directly into the AWS dashboard, allowing customers to deploy the model using existing credits and infrastructure.

But the partnership goes beyond simple integration. One of the most significant hurdles for enterprise AI is data privacy. Companies are often unwilling to move sensitive data to third-party infrastructure. To solve this, Fundamental and Amazon achieved a massive engineering feat: the ability to deploy fully encrypted models—both the architecture and the weights—directly within the customer’s own environment.

“Customers can be confident the data sits with them,” Fraenkel explains. “We are the first, and currently only, company to have built such a solution.”

The Spreadsheet Finally Gets Its AI Moment

Fundamental’s emergence represents more than just another AI tool—it’s an attempt to redefine the operating system for business decisions. If NEXUS performs as advertised—handling financial fraud, energy prices, and supply chain disruptions with a single, generalized model—it will mark the moment where AI finally learned to read the spreadsheets that actually run the world.

The Power to Predict is no longer about looking at what happened yesterday; it’s about uncovering the hidden language of tables to determine what happens tomorrow.

Tags & Viral Phrases:

• The spreadsheet revolution
• AI finally cracks business data
• Large Tabular Model (LTM)
• DeepMind alumni startup
• $255 million funding bombshell
• Enterprise AI transformation
• The hidden language of tables
• Order-invariant AI architecture
• Predictive intelligence as utility
• Single line of code prediction
• Fortune 100 contracts secured
• AWS Marketplace gamechanger
• Data privacy solved
• Beyond the bottom line AI
• Infrastructure safety predictions
• Healthcare optimization AI
• Climate resilience forecasting
• Patient outcomes AI
• The spreadsheet that runs the world
• Business decisions OS
• What happens tomorrow
• The future of enterprise AI
• Tabular data revolution
• Numbers problem solved
• Sequential logic vs. tabular logic
• Manual feature engineering eliminated
• Amazon SageMaker HyperPod
• Encrypted model deployment
• AWS credits for AI
• Python-based predictive interface
• Silent high-speed decision engine
• Marginal accuracy massive value
• The $225 million vote of confidence
• Annie Lamont Oak HCFT
• Salesforce Ventures investment
• Valor Equity Partners
• Battery Ventures
• Perplexity leaders investment
• Wiz Brex Datadog angels
• The DeepMind mafia strikes again
• Enterprise data privacy breakthrough
• Single mother ER prediction
• Flint water crisis prevention
• COVID PPE shortage prediction
• Pakistan floods climate prediction
• Hospital readmission AI
• Credit default probability
• Maintenance risk score
• Equipment failure prediction
• Fraud detection in real-time
• Customer churn prediction
• Transaction frequency analysis
• Support ticket sentiment
• Regional economic conditions
• Tokenization trap solved
• Numbers understood natively
• Non-linear interaction detection
• The future is tabular
• AI that reads spreadsheets
• Business intelligence reimagined
• The spreadsheet gets smart
• Enterprise AI foundation model
• The next frontier in AI
• Structured data deep learning
• The missing piece of AI

,