Fundamental Raises $255 Million to Launch NEXUS: AI Tailored for Enterprise Spreadsheets and Tabular Data

In the world of business, decisions often boil down to numbers in rows and columns. Spreadsheets, databases, and structured records form the backbone of enterprise operations, capturing everything from pricing strategies and sales forecasts to risk assessments and customer patterns. Yet, while artificial intelligence has made massive strides in handling text, images, and videos, it has largely overlooked this critical realm of tabular data. This oversight leaves companies stuck with legacy tools for predictions that shape their bottom lines.

Enter Fundamental, a San Francisco-based AI startup that’s addressing this gap. Founded by former DeepMind researchers, the company has raised a whopping $255 million in funding and officially launched NEXUS, its flagship Large Tabular Model (LTM). Designed specifically to predict outcomes from enterprise spreadsheets and structured data, NEXUS promises to bridge this gap, enabling smarter, faster business decisions without the usual headaches of data preparation.

This launch isn’t just about new tech—it’s a response to a real pain point. Most organizations still rely on manual processes or basic statistical models to crunch their tabular data, which can be time-consuming and error-prone. NEXUS aims to automate that, turning raw tables into actionable insights with minimal intervention.

The Overlooked Power of Tabular Data in Business

Tabular data isn’t glamorous like chatbots or image generators, but it’s the lifeblood of enterprises. Think about it: pricing models in retail, inventory forecasts in manufacturing, fraud detection in finance, or patient outcomes in healthcare—all powered by spreadsheets and databases. These structures hold the world’s most valuable business intelligence, yet they’ve been sidelined by the AI boom.

Why the neglect? Traditional AI models, like large language models (LLMs), excel at unstructured data but falter with tables. LLMs might summarize a report or generate code, but parsing complex interactions between rows and columns—such as how a customer’s purchase history correlates with seasonal trends—requires specialized handling. Without it, businesses miss out on predictive power that could optimize operations and drive revenue.

The consequences are clear. Companies waste hours on feature engineering, where data scientists manually tweak variables to fit models. This slows down decision-making in fast-paced markets. For instance, a logistics firm might delay route optimizations because their spreadsheet tools can’t scale to millions of records. Or a financial team could overlook subtle risk patterns in transaction data, leading to costly oversights.

Fundamental sees this as an opportunity. By focusing on AI for enterprise spreadsheets, they’re targeting a market ripe for innovation. Their approach builds on the idea that tabular data deserves its own foundation models, much like how LLMs were built for language. This shift could generate trillions in value, as tables underpin decisions across every industry vertical.

Fundamental: From DeepMind Roots to AI Innovation

Fundamental was founded in October 2024 by a team of ex-DeepMind researchers, including CEO and co-founder Jeremy Fraenkel. Emerging from stealth mode, the startup is laser-focused on predictive analytics for structured data. Unlike general-purpose AI firms, Fundamental’s mission is narrow but deep: help enterprises forecast what happens next, rather than just dissecting the past.

“We’ve built a generalised foundation model specifically to leverage the world’s most valuable data: the billions of tables that underpin predictions in every enterprise, across every vertical,” says Fraenkel. “NEXUS is the OS for business decisions.”

This “operating system” analogy makes sense. Just as an OS manages hardware resources, NEXUS orchestrates tabular data flows, making predictions accessible to non-experts. The team’s DeepMind background brings credibility—DeepMind is renowned for breakthroughs in reinforcement learning and protein folding, skills that translate well to modeling complex dependencies in data tables.

Since its inception, Fundamental has prioritized practical applications. They’ve assembled a research team blending academic rigor with commercial savvy, ensuring their tech isn’t just theoretical. Early traction shows promise: the company has already inked seven-figure contracts with Fortune 100 companies, deploying NEXUS for real-world predictions like demand forecasting and operational risk analysis.

Breaking Down the $255 Million Funding Round

The launch of NEXUS is fueled by substantial backing. Fundamental’s funding totals $255 million, comprising a $30 million seed round and a massive $225 million Series A. The Series A was led by Oak HC/FT, a venture firm with a track record in health tech and fintech, signaling strong belief in NEXUS’s cross-industry potential.

Other participants include:

• Valor Equity Partners, known for scaling enterprise software.
• Battery Ventures, experts in AI and cloud infrastructure.
• Salesforce Ventures, bringing CRM and data integration know-how.
• Hetz Ventures, focused on deep tech innovations.

This investor lineup isn’t random. It reflects NEXUS’s appeal to sectors craving better data tools, from e-commerce to energy. The capital injection will support key growth areas:

• Scaling compute resources: Training LTMs requires immense processing power, and Fundamental plans to ramp up GPU clusters for faster model iterations.
• Expanding enterprise deployments: More pilots and integrations to prove ROI in diverse settings.
• Team growth: Hiring in research, engineering, and go-to-market functions to handle demand.

With this war chest, Fundamental is positioned to compete in the burgeoning predictive AI for business space. While exact valuations aren’t public, the round size suggests a valuation north of $1 billion, putting them in unicorn territory early on.

Funding Stage	Amount	Lead Investor	Key Participants
Seed Round	$30 million	N/A	Early backers
Series A	$225 million	Oak HC/FT	Valor Equity Partners, Battery Ventures, Salesforce Ventures, Hetz Ventures

This table highlights the structured support behind Fundamental, mirroring the tabular data their tech handles.

NEXUS: A New Era for Large Tabular Models

At the heart of Fundamental’s offering is NEXUS, billed as the most powerful Large Tabular Model to date. Unlike LLMs trained on internet text, NEXUS is purpose-built for tables. It’s trained from scratch on billions of tabular datasets, learning the nuances of how data points interconnect—think correlations between variables that humans might miss.

What sets NEXUS apart?

• Native Table Understanding: It processes rows, columns, and relationships without converting data into text first, avoiding the inefficiencies of LLMs.
• No Heavy Setup Required: Enterprises integrate it via a single line of code into existing stacks like SQL databases or cloud warehouses.
• Auto-Learning Capabilities: Once connected, NEXUS ingests raw data and detects patterns, dependencies, and structures automatically. No need for extensive feature engineering or custom training.

This ease-of-use addresses a major need for busy teams. Imagine a marketing analyst uploading customer behavior spreadsheets; NEXUS could predict churn rates or campaign ROI in minutes, not weeks.

“The significance of Fundamental’s model is hard to overstate – structured, relational data has yet to see the benefits of the deep learning revolution,” said Annie Lamont, Co-Founder & Managing Partner at Oak HC/FT. “Fundamental’s ability to predict anything from financial fraud to hospital readmission to energy prices positions the company to support virtually every industry and sector. With a world-class research team that blends deep technical expertise with proven commercial execution, the company brings a rare mix of research rigor and enterprise GTM understanding. We’re honored to be partnering with them on their journey.”

Lamont’s words underscore NEXUS’s versatility. In finance, it could flag fraudulent transactions by spotting anomalies in ledger tables. In healthcare, analyzing patient records to forecast readmissions. In energy, modeling price fluctuations from supply chain data. The model’s scalability handles enterprise volumes—terabytes of structured info—without breaking a sweat.

Compared to traditional tools like Excel’s forecasting functions or basic machine learning libraries, NEXUS offers depth. It doesn’t just fit lines to data; it builds contextual understanding, improving accuracy over time as it learns from feedback loops.

Seamless Integration and Real-World Adoption

Adoption is where NEXUS shines. Fundamental emphasizes plug-and-play deployment, compatible with popular data ecosystems. Users connect their spreadsheets or databases, and the model does the rest: cleaning, structuring, and predicting.

Key benefits for enterprises include:

1. Reduced Time to Insight: Skip the weeks of data prep; get predictions in hours.
2. Cost Savings: Less reliance on data scientists for every query.
3. Enhanced Accuracy: By focusing on tabular specifics, NEXUS outperforms generic AI on structured tasks.
4. Security and Compliance: Built with enterprise-grade protections, handling sensitive data like financial records or health info.

Early adopters—those Fortune 100 firms—report using NEXUS for high-stakes work. One example might involve supply chain optimization, where predicting disruptions from vendor tables prevents stockouts. Another could be sales forecasting, integrating CRM data with historical trends for precise quotas.

Fundamental claims NEXUS establishes a new AI category, one centered on tables rather than text or visuals. This makes sense in a data-driven economy, where 80% of enterprise data is structured yet underutilized. By democratizing predictions, NEXUS empowers analysts, executives, and even frontline workers to make data-backed calls.

Strategic Partnership with AWS: Accelerating Accessibility

To supercharge reach, Fundamental announced a partnership with Amazon Web Services (AWS). This collaboration lets AWS customers access NEXUS directly from their dashboards, alongside services like EC2 or S3. It’s as simple as selecting a model and deploying—no custom infrastructure needed.

“Fundamental’s structured data prediction model builds on AWS’s advanced AI offerings, helping enterprise customers fill a crucial gap in comprehensive tabular data analysis at scale,” said Dave Brown, VP of Compute, Platforms & ML Services at AWS. “By partnering with Fundamental, we are making it seamless for customers to transform tabular data – the backbone of enterprise decision-making – into a powerful predictive asset. This collaboration exemplifies our commitment to bringing transformative AI solutions to market with the enterprise-grade security and scalability our customers demand.”

This tie-up is strategic. AWS dominates cloud AI, serving millions of businesses. Integrating NEXUS here exposes it to a vast audience, from startups tweaking spreadsheets to giants managing petabyte-scale databases. It also ensures compliance with standards like GDPR and HIPAA, crucial for tabular data often containing personal or proprietary info.

For users, the benefits are tangible:

• One-Stop Shop: Manage all AI workloads in AWS.
• Scalable Pricing: Pay-as-you-go for predictions, aligning with usage.
• Hybrid Support: Works with on-prem data if needed, easing migrations.

This partnership could accelerate NEXUS’s market penetration, much like how AWS boosted early adopters of tools like SageMaker.

Industry Applications: Unlocking Predictions Across Sectors

NEXUS’s design lends itself to broad applications. Let’s explore how it fits into key industries, drawing on its core strengths in AI for enterprise spreadsheets.

Finance and Risk Management

In banking, tabular data rules: transaction logs, credit scores, market feeds. NEXUS can predict fraud by analyzing patterns in real-time tables, reducing false positives that plague rule-based systems. For risk, it models portfolio exposures from historical trades, helping traders anticipate downturns.

Retail and E-Commerce

Customer behavior tables—purchases, browses, returns—are goldmines. NEXUS forecasts demand, personalizes recommendations, and optimizes pricing dynamically. A retailer could use it to predict holiday surges, adjusting inventory without overstocking.

Healthcare and Life Sciences

Patient records, trial data, and genomic tables are highly structured. Here, NEXUS shines in predicting outcomes like readmissions or drug efficacy, supporting evidence-based care. Hospitals might integrate it with EHR systems for proactive interventions.

Manufacturing and Supply Chain

From production schedules to supplier metrics, tables track operations. NEXUS predicts maintenance needs or supply disruptions, minimizing downtime. In volatile markets, it could forecast raw material costs from global trade data.

Energy and Utilities

Price volatility and grid data are tabular nightmares. NEXUS models consumption patterns, aiding renewable integration or outage predictions. Utilities benefit from accurate load balancing, cutting energy waste.

Across these, the common thread is prediction at scale. NEXUS doesn’t replace human judgment but augments it, turning static spreadsheets into dynamic tools.

Challenges and the Road Ahead for Tabular AI

No tech is without hurdles. Training on billions of datasets raises questions about data sourcing—Fundamental ensures ethical, anonymized training to avoid biases. Scalability is another: while AWS helps, enterprises with legacy systems might face integration snags.

Looking forward, Fundamental’s plans with the new funding are ambitious. They’ll expand research into multimodal tables (e.g., blending spreadsheets with text notes) and refine NEXUS for edge cases like sparse data. Team growth will bolster sales, targeting mid-market firms beyond Fortune 100.

The broader Large Tabular Model landscape is nascent but growing. Competitors might emerge, but Fundamental’s first-mover edge and DeepMind pedigree give them a head start. As AI matures, expect more focus on structured data, potentially integrating LTMs with LLMs for hybrid analytics.

Why NEXUS Matters for the Future of Business Intelligence

Fundamental’s launch of NEXUS with $255 million in tow marks a pivotal moment for AI for enterprise spreadsheets. By tackling the tabular data gap, they’re equipping businesses to predict with precision, driving efficiency and innovation. Whether you’re a CFO eyeing risks or a supply chain manager planning ahead, tools like NEXUS could redefine how we work with data.

As enterprises digitize further, the demand for such specialized AI will only grow. Fundamental isn’t just building a model; they’re providing the groundwork for a more predictive economy. If their early successes hold, NEXUS could become as essential as spreadsheets themselves—only smarter.