SAP said on 4 May 2026 that it had agreed to acquire Dremio, a US data lakehouse company whose software lets people and AI agents query company data stored in open formats without first copying it into a warehouse. SAP said it was buying all of the company.
Terms were not disclosed when the deal was announced. SAP's half-year report, filed with the US Securities and Exchange Commission, later said the deal closed on 6 July 2026 and that the cash paid amounted to about €0.5 billion (about $580 million), which it described as the main part of the consideration.
SAP plans to use Dremio to turn SAP Business Data Cloud into a lakehouse built natively on Apache Iceberg, the open table format, so that SAP and non-SAP data can sit side by side without being moved or converted. Dremio's leaders wrote that the company "isn't going anywhere" and that customers would gain from SAP's scale.
What Dremio built
Dremio was started in 2015 by Tomer Shiran and Jacques Nadeau, who had both just left the big data company MapR. Redpoint Ventures backed them early, and the pair worked out of Redpoint's Menlo Park offices, investor Tomasz Tunguz wrote in 2020.
The founders set out to build a fast query engine for data lakes, which at the time were used mostly by data scientists. Dremio co-created Apache Arrow, an in-memory data format, and built its SQL engine on it. It added fast federated queries across other databases and a caching technology called Reflections. In early 2021 it adopted Apache Iceberg, rebuilt its engine around the format, and later co-created the Apache Polaris catalog with Snowflake.
More recently Dremio repositioned itself as an "agentic lakehouse". It says it shipped an MCP interface so AI agents can query data directly, added an AI semantic layer for business context, and automated table maintenance tasks such as clustering and compaction. It names Shell, TD Bank and Michelin among its customers. Sendur Sellakumar was chief executive at the time of the deal.
“Enterprise AI doesn’t stall because the models aren’t good enough; it stalls because the data isn’t ready for AI agents.”
Why SAP is buying
SAP's pitch is that enterprise AI projects stall because data is scattered, locked in proprietary formats and stripped of business context. Philipp Herzig, SAP's chief technology officer, said Dremio removed that bottleneck and that, combined with Business Data Cloud, SAP could take customers from fragmented data to governed, AI-ready data on one open platform.
SAP said Dremio will provide an open catalog built on Apache Polaris and the Apache Iceberg REST Catalog API. The catalog is meant to act as the discovery and semantic layer of Business Data Cloud and the base of the SAP Knowledge Graph. SAP said it was committed to keep investing in Iceberg, Polaris and Arrow. SAP chief executive Christian Klein had told analysts on the first-quarter earnings call that AI agents need to work with many types of data, Constellation Research noted.
The funding
Dremio raised a $70 million Series C in 2020 and a $135 million Series D, led by Sapphire Ventures, a little over a year before its next round. In January 2022 it raised a $160 million Series E led by Adams Street Partners that valued the company at $2 billion, double its previous valuation. Sapphire Ventures, Insight Partners, Lightspeed Venture Partners, Norwest Venture Partners and Cisco Investments took part, and StepStone Group and DTCP joined as new investors.
Dremio was based in Santa Clara, California, at the time of the Series E. The deal announcement was datelined Walldorf and Austin. Neither company disclosed Dremio's headcount.
Mixed reviews from analysts
Holger Mueller of Constellation Research said the deal showed how quickly SAP's data setup for AI was changing. He also described Dremio as commercially struggling, said the strategy relies on zero-copy data federation that has not yet proven itself at scale, and warned that customers could find it confusing because SAP already partners with Databricks on Business Data Cloud. His colleague Mike Ni said Dremio pushed Business Data Cloud toward a more open foundation for SAP and non-SAP data.
SAP announced a second acquisition the same day: Prior Labs, a German maker of AI models for tabular data. SAP said the two deals, both closed in July, would reduce its 2026 operating profit by more than €100 million (about $117 million).
