Analyst(s): Brad Shimmin
Publication Date: May 6, 2026
SAP has announced its intent to acquire Dremio and Prior Labs to unify heterogeneous enterprise data and accelerate the development of agentic AI. By integrating Dremio’s Apache Iceberg-native lakehouse and Prior Labs’ tabular foundation models, SAP aims to eliminate the friction between SAP and non-SAP data sources. This move represents a significant investment in the infrastructure required to power autonomous software agents with deeply contextualized business information.
What is Covered in This Article:
- SAP’s acquisition of Dremio and its integration into the SAP Business Data Cloud to provide a federated, Apache Iceberg-native environment.
- The strategic role of Prior Labs and SAP’s €1 billion commitment to developing Tabular Foundation Models (TFMs) for structured business data.
- The technical synergy between open data catalogs like Apache Polaris and the SAP Knowledge Graph.
- How the unification of infrastructure and intelligence layers enables the transition from predictive analytics to autonomous agentic AI.
- Market implications for enterprise data governance and the shift toward open, federated architectures.
The News: On May 4, 2026, SAP announced two pivotal acquisitions designed to bridge the gap between fragmented enterprise data and the next generation of artificial intelligence. SAP has entered into an agreement to acquire Dremio, a leading independent data lakehouse company, and Prior Labs, a specialized AI laboratory focused on tabular data. The acquisition of Dremio brings a serverless, open-standard architecture—centered on Apache Iceberg—directly into the SAP Business Data Cloud. This integration enables organizations to query SAP and third-party data without the costly, complex data movement required by traditional approaches.
Simultaneously, SAP has committed to a €1 billion investment over four years into Prior Labs, which will operate as an independent entity to advance Tabular Foundation Models (TFMs). These models specialize in interpreting the structured relational data—such as financial ledgers and inventory records—that form the backbone of corporate operations. The Dremio transaction is slated to close in the third quarter of 2026, while the Prior Labs acquisition is expected to finalize in the second or third quarter of 2026, both subject to customary regulatory approvals.
Can SAP Finally Kill the ETL Monster? The Dremio and Prior Labs Acquisition Explained
Analyst Take: The persistent headache for any Chief Data Officer has always been the data tax—the enormous amount of time, money, and compute power wasted on moving data from where it lives to where it can be used. By acquiring Dremio and Prior Labs, SAP is attempting to repeal that tax rather than simply adding features to a legacy stack. For decades, the enterprise has been divided into the structured, highly governed world of ERP and the decentralized sprawl of the cloud data lake. By bridging these environments through open standards like Apache Iceberg, SAP is making a pragmatic bet that the future of the intelligent enterprise depends on how easily an AI can access data it does not natively own.
Breaking the Proprietary Gravity of Enterprise Data
The technical significance of Dremio in the SAP ecosystem lies in its decoupling of storage and compute. Historically, analyzing non-SAP data alongside S/4HANA records required fragile ETL (extract, transform, load) pipelines. This process often resulted in data becoming stale before it ever reached a dashboard. Dremio changes this dynamic by using Apache Iceberg as its native table format. Iceberg acts as a universal translator, allowing different engines to read and write the same data concurrently without corruption or redundant data duplication.
According to the 1H 2026 DIAI Market Sizing & Five-Year Forecast Report, the demand for federated, open-standard data platforms continues to accelerate as enterprises seek to reduce the overhead associated with large-scale data governance. By adopting an Iceberg-first strategy, the SAP Business Data Cloud effectively transforms into an open environment. Organizations no longer have to migrate third-party operational data into proprietary SAP structures to generate value. These assets can stay in their original cloud storage buckets, while Dremio provides the high-performance query layer that makes them appear native to the SAP environment. Mitigating data movement in this way represents a massive win for architectural flexibility.
The Strategic Impact of SAP Acquiring Dremio and Prior Labs
While Dremio solves the problem of where data lives, Prior Labs addresses what that data actually means. Despite the excitement surrounding Large Language Models (LLMs), enterprise leaders are realizing that an LLM often represents the wrong tool for critical business tasks involving structured information. Standard LLMs typically fail to navigate the mathematical precision and statistical relationships inherent in relational records. Businesses run on tables, not just text.
Prior Labs specializes in Tabular Foundation Models (TFMs), including its TabPFN-2.6 model. These are statistical reasoning engines that learn directly from tabular data rather than functioning as general-purpose chatbots. Benchmark results on TabArena indicate that TabPFN can match the accuracy of extensive automated machine learning pipelines almost instantly. By investing €1 billion into this technology, SAP acknowledges that the intelligence in artificial intelligence must be as structured as the business data it analyzes. This allows SAP to offer zero-shot predictions and in-context learning, enabling models to provide insights on enterprise data without requiring constant, manual retraining. Importantly, this concept is not new to SAP, which has already invested in TFMs with Relational Pretrained Transformer-1 (RPT-1), an SAP-built model that’s already generally available within its Business Technology Platform.
From Knowledge Graphs to Agentic Reality
The real magic happens when you connect these two acquisitions through the SAP Knowledge Graph. The industry is currently moving toward an agentic model, where AI systems like SAP’s Joule don’t just answer questions but autonomously reason through problems and execute actions. For an agent to be useful, it needs a semantic map of the business. The SAP Knowledge Graph provides that map, embedding business relationships, organizational hierarchies, and regulatory classifications directly into the data layer.
Dremio feeds this graph with raw, federated data, and Prior Labs provides the reasoning engine required to understand causality within business metrics. This allows an AI agent to see that a supply chain disruption in a federated non-SAP system will impact an inventory count in SAP S/4HANA, and then use a TFM to predict the resulting operational risk. The agent can then autonomously initiate workflows to mitigate that risk. This level of autonomy remains impossible as long as data is trapped in silos or AI models fail to natively comprehend the statistical relationships of the enterprise.
The Open Catalog: Why Apache Polaris Matters
A critical detail in this news is the role of Apache Polaris. Co-created by Dremio and Snowflake, Polaris is an open-source catalog designed to govern Iceberg tables across diverse environments. By delivering a universal catalog based on Apache Polaris and the Apache Iceberg REST Catalog API, SAP is signaling a commitment to being a responsible citizen in the open data ecosystem. This marks a departure from the closed data siloes of the past. It suggests that SAP is comfortable in a world where customers use a variety of different tools, provided that SAP remains the central nervous system that coordinates them.
The pragmatic reality is that no enterprise is 100% SAP. Modern businesses are collections of different clouds, databases, and specialized analytics clusters. The integration of Dremio and Prior Labs, which follows the March 2026 acquisition of master data management specialist Reltio, is a refreshingly honest admission of this fact. It provides a way for SAP to extend its governance and intelligence over data it doesn’t directly control, which ultimately makes the SAP platform more indispensable by serving as the foundation for the agentic enterprise.
What to Watch:
- While the Dremio deal is slated to close in Q3 2026, the real test will be how quickly Dremio’s federated query capabilities become a native feature within SAP Datasphere and the broader Business Data Cloud.
- SAP is maintaining Prior Labs as an independent entity in Freiburg. Enterprise practitioners should watch how they balance academic research with the need to scale the TabPFN models for immediate commercial use within SAP’s Joule agents.
- Since Dremio offers a high-performance alternative to some native cloud query services, look for how AWS, Azure, and Google Cloud adjust their data partnership strategies with SAP.
- As SAP doubles down on the “fourth layer” of its strategy (i.e., bringing AI to structured data), expect other major enterprise players, such as Oracle and Salesforce, to accelerate their own research or acquisitions in the Tabular Foundation Model space.
See the complete press release on SAP’s intent to acquire Dremio to power agentic AI on the SAP website.
Disclosure: Futurum is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of Futurum as a whole.
Other Insights From Futurum:
Can Databricks Out-Iceberg the Competition?
Author Information
Brad Shimmin is Vice President and Practice Lead, Data Intelligence, Analytics, & Infrastructure at Futurum. He provides strategic direction and market analysis to help organizations maximize their investments in data and analytics. Currently, Brad is focused on helping companies establish an AI-first data strategy.
With over 30 years of experience in enterprise IT and emerging technologies, Brad is a distinguished thought leader specializing in data, analytics, artificial intelligence, and enterprise software development. Consulting with Fortune 100 vendors, Brad specializes in industry thought leadership, worldwide market analysis, client development, and strategic advisory services.
Brad earned his Bachelor of Arts from Utah State University, where he graduated Magna Cum Laude. Brad lives in Longmeadow, MA, with his beautiful wife and far too many LEGO sets.
