SAP is moving to fix a problem that has quietly held back enterprise AI. The company is acquiring, targeting two weak points that most organizations still struggle with: fragmented data and poor AI performance on structured datasets.
At its core, the problem is simple. Enterprise AI struggles because data is often not accessible in real time across systems, and the models are not built for the structured data that drives most business decisions.
The acquisitions reflect a shift in how SAP approaches this problem altogether. Instead of adding features on top of existing systems, the company is trying to rebuild the foundation. It is connecting how data is accessed with how models actually work on it. And if it succeeds, SAP moves from being a system of record to something closer to an AI execution layer across the enterprise.
To understand how the two companies fit into SAP’s strategy, let’s dig into what they do. Dremio focuses on access. It allows data to be queried across systems without moving it first, so SAP and non SAP data can be used together in real time. Prior Labs focuses on how that data is used. It builds models designed for structured datasets like finance, supply chain, and operations, where most enterprise decisions sit.
The language SAP is using leaves little ambiguity. The focus on “unify SAP and non SAP data to power agentic AI,” alongside its push for “a unified, open data layer,” reflects a break from its historical model. Data is no longer expected to sit inside SAP. It already exists across cloud platforms and external systems.
What changes is SAP’s position. It moves from owning the data environment to operating across it. That places it closer to the access layer, which is where AI systems depend on control. In this model, value comes from connecting data, not containing it. SAP is shifting accordingly.
SAP is not being subtle about the constraint. “Enterprise AI doesn’t stall because the models aren’t good enough; it stalls because the data isn’t ready for AI agents,” said Philipp Herzig, CTO, SAP SE. The follow up is just as direct. “Dremio eliminates that bottleneck.”
That positioning connects directly to SAP’s broader platform play. By tying Dremio to SAP Business Data Cloud and a “single open platform,” the company is not just identifying the bottleneck. It is defining how to remove it. Data is no longer something that needs to be prepared before AI can run. It becomes part of the system AI operates on directly.
Access does not solve everything. Even with the right data, most models are not built to handle structured datasets. That is where enterprise AI still breaks. SAP’s acquisition of Prior Labs is aimed at fixing that.
The company builds models for tabular data rather than text, making them better suited for business environments. Instead of forcing structured data into formats designed for language models, the models operate on it directly.
For SAP, this completes the setup it is building. Data can be accessed across systems, and the models can work on that data without reshaping or delay. Both pieces are needed for AI to function in production.
“Early on, SAP recognized that the greatest untapped opportunity in enterprise AI wasn’t large language models; it was AI built for the structured data that runs the world’s businesses,” SAP CTO Philipp Herzig said.
“We built SAP-RPT-1 to prove that conviction for enterprise data. Prior Labs has built a leading TFM on public benchmarks and built one of the leading research teams in this category. Combining their frontier model work with enterprise data and customer reach is how we intend to lead this category globally.”
With both pieces in place, SAP moves closer to data platforms that have already been building similar stacks. Databricks and Snowflake have spent years combining data access and AI into unified systems, while hyperscalers are pushing agent-based workflows across their clouds. SAP starts from a different position. It operates where enterprise data already lives. That gives it an advantage in context, however, it also needs to match the speed and openness of platforms built for this from the start.
What SAP is building goes beyond its traditional stack. It is trying to control how data turns into decisions. That creates both opportunity and risk. Bringing open data access into SAP’s ecosystem, aligning it with existing systems, and deploying new model approaches across workflows are not small changes. They require speed and coordination. It may also require a change in how SAP operates. While the strategy is clear, the challenge is execution, especially in a market where competitors are already moving fast.
Editor’s note: This article originally appeared in BigDATAwire.
The post SAP Seeks Solution to Enterprise AI Puzzle with Dremio, Prior Labs Acquisitions appeared first on AIwire.
