SAP announced moves to strengthen the foundations of enterprise AI with planned acquisitions of data platform provider Dremio and AI research firm Prior Labs, aiming to address two persistent constraints: fragmented data environments and limited model performance on structured business data. The two May 4 news releases signal a more deliberate push to align data infrastructure and AI model development as a single architectural priority rather than parallel initiatives.
SAP CEO Christian Klein framed the moves as part of a unified strategy to turn enterprise data into actionable intelligence. He said the acquisitions build on SAP’s existing data foundation to help customers “turn data into trusted decisions and predictive insights.”
Unifying Enterprise Data with Dremio
SAP’s planned acquisition of Dremio focuses on the data layer, where fragmentation and integration challenges continue to limit AI adoption.
According to SAP, many enterprise AI initiatives fail not because of model limitations but because underlying data is “fragmented, locked in proprietary formats and stripped of the business context that makes it meaningful.” This results in pilots that cannot scale, duplicated engineering effort, and increased compliance risk.
Dremio’s lakehouse platform is designed to address these issues by enabling unified access to SAP and non-SAP data without requiring data movement or format conversion. Once integrated, SAP Business Data Cloud will become an Apache Iceberg-native environment, allowing structured and unstructured data to coexist on a single open foundation.
“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. “Dremio eliminates that bottleneck.”
The platform will also introduce a universal data catalog layer, combining metadata, lineage, and access controls into a single semantic framework. This is expected to form the foundation of SAP’s Knowledge Graph, embedding business context directly into data used for analytics and AI workloads.
Analysis
What this means: Open data architectures are moving from optional to foundational. Federated access, lakehouse models, and open formats are becoming prerequisites for scaling AI across heterogeneous environments. For enterprise architects and data leaders, this raises the bar on governance, data quality, and interoperability as core design requirements, especially in multi-vendor ERP landscapes.
Expanding AI Capabilities with Prior Labs
While Dremio addresses data readiness, SAP’s planned acquisition of Prior Labs targets the model layer, specifically AI performance on structured enterprise data.
LLMs have struggled with tabular data, which dominates enterprise systems such as finance, supply chain, and procurement. Prior Labs specializes in Tabular Foundation Models (TFMs), which are designed to understand and predict outcomes from structured datasets such as payment behavior, supplier risk, and customer churn.
SAP plans to invest more than €1 billion (approximately $1.17 billion) over four years to scale Prior Labs into a global frontier AI lab focused on this category. The unit will operate independently to maintain research velocity while integrating with SAP’s product stack through SAP AI Core, SAP Business Data Cloud, and the Joule agent layer.
“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,” Herzig said.
Prior Labs’ technology, including its widely adopted open-source model TabPFN, enables instant predictions on tabular data without requiring extensive model training. SAP said this capability will allow business users to run “what-if” scenarios and generate predictions directly from enterprise datasets, with minimal data science expertise.
Analysis
What this means: Structured enterprise data is becoming the primary source of AI differentiation. While much of the AI conversation has centered on large language models, most ERP value sits in structured datasets across finance, supply chain, and operations. Vendors that improve prediction, simulation, and decision-making directly on that data will define the next phase of competitive advantage for ERP buyers.
Sponsor Industry‑Grade Research
Aligning Data Infrastructure and AI Execution
The two acquisitions reflect a broader architectural shift. Dremio extends SAP’s ability to unify and govern enterprise data across systems, while Prior Labs enhances its ability to generate predictive insights from that data.
Klein described this as building on SAP’s “strong data foundation,” combining open data infrastructure with advanced models tailored to business data.
The strategy also aligns with SAP’s broader push toward agentic AI, where systems not only analyze data but execute workflows and decisions autonomously. In that context, both data accessibility and model accuracy become critical prerequisites.
SAP said the combined capabilities will allow customers to move from “raw, fragmented data to governed, AI-ready intelligence” while improving time to value for AI initiatives.
Analysis
What this means: ERP platforms are converging on full-stack AI architectures. This move reinforces a broader market shift where vendors are investing across data, models, and applications simultaneously. For ERP leaders, the implication is that evaluating AI capability will increasingly require assessing how well vendors integrate these layers into a cohesive system, not just the surface-level features delivered in applications.
