We at Fundamental are partnering with SAP to bring our most powerful Large Tabular Model NEXUS into the SAP Business AI platform. NEXUS will soon be available to SAP customers through SAP's open model ecosystem.
Why does this matter? Tabular data runs the enterprise
Most of the AI conversation over the last three years has been about Large Language Models. That's understandable: LLMs are remarkable, and they've changed how millions of people work. They do an astonishing job of processing unstructured data like images, video and text.
But if you look at how decisions actually get made inside a large company - where revenue is forecasted, where risk is priced, where supply chains are planned, where churn is predicted, where fraud is caught - you're not looking at unstructured data. You're looking at tables: millions of rows and columns across vast ERP systems, CRMs, data warehouses, and operational databases.
This is the data that runs the global economy. And, until very recently, the only way to get predictions out of it was to hire a team of data scientists, spend months building a bespoke model for each use case, and retrain it constantly. Most companies never got past the first few use cases.
NEXUS changes that. It's a Large Tabular Model (LTM) trained on tables in the same way that LLMs are trained on text. You point our model at a table, you tell it what you want to predict, and it predicts. No bespoke model. No months of feature engineering. No army of data scientists.
Our partnership philosophy
Our philosophy at Fundamental is straightforward: bring the model to the data, not the data to the model.
Enterprise data lives in many places, and customers shouldn't have to move it, copy it, or redesign their workflows to experience the benefits of NEXUS. That's why we're committed to making NEXUS available everywhere our customers' data already sits, whether that's inside the major hyperscaler platforms, or within enterprise application providers like SAP.
The SAP partnership reflects this philosophy in action. SAP is enabling an open model ecosystem through its generative AI Hub, allowing customers the flexibility to select and switch between models, including NEXUS, without redesigning their workflows. We expect to announce similar availability across other platforms in the months ahead.
"Enterprises run on structured business data, and tabular AI is critical to unlocking its full value," said Jonathan von Rueden, Chief AI Officer of SAP SE. "We're excited to add Fundamental’s NEXUS, a leading Large Tabular Model, to our family of best-in-class models. By advancing large tabular models, we can move beyond historical analysis toward adaptive, high‑quality predictions.”
What this unlocks for enterprises
For SAP customers, the practical impact is simple and significant. The kinds of prediction problems that used to require a custom modeling project such as demand forecasting, credit risk, and predictive maintenance can be addressed by pointing NEXUS at the relevant data and "and getting a working model without months of feature engineering.
This changes the fundamental economics of enterprise AI: the time-to-value reduces significantly and the bar for which problems are worth solving with AI drops considerably, which means a lot more of them get solved.
The bigger picture
We started Fundamental because we believed the same scaling story that played out in unstructured data with LLMs was going to play out in structured data with LTMs. We're building the foundation model that will define how enterprises work with their structured data, and we're going to make it available wherever that data lives. Today's news is one step toward that, and there will be many more.
