AWS buys the company behind DuckDB
Amazon Web Services (AWS) has signed a definitive agreement to acquire DuckLabs, the Amsterdam-based company behind the open source analytics database DuckDB. The deal was announced August 26, 2026, with financial terms undisclosed by either party. AWS says the transaction is expected to close subject to customary conditions, without giving a specific date.
Why DuckDB matters
DuckDB is an embeddable, in-process analytical database, often described as "SQLite for analytics" because it runs inside an application rather than as a separate server process. It has become one of the most widely adopted open source data tools of the past few years, with peak usage reported at more than 3 million downloads per day. DuckLabs co-founders Hannes Muhleisen and Mark Raasveldt, who originally created DuckDB, will join AWS along with the rest of the Amsterdam engineering team and continue leading the project's technical direction.
The open source promise
AWS is acquiring the commercial company, not the open source project itself. DuckDB will remain free software under the MIT license and will keep being governed independently by the nonprofit DuckDB Foundation, according to AWS's announcement. AWS also committed to continued support for two related open source projects, DuckLake and Quack. Constellation Research analyst Michael Ni told SiliconANGLE that AWS is "smartly keeping DuckDB open source, so it can continue to spread across the broader ecosystem," a strategy that gives AWS influence over where analytical workloads run without having to control the software's license.
What "in-process" actually buys you
DuckDB's in-process design means it runs as a library inside the application doing the analysis, a notebook, a script, a backend service, rather than as a separate server a client connects to over a network. That is what earns it the "SQLite for analytics" comparison: no cluster to provision, no server to keep running, just a fast query engine linked directly into the calling process. For AWS, embedding that kind of engine close to where data already sits in S3 is a more direct route to letting customers query in place than standing up a full separate data warehouse for every analytical job.
Not a cold start
AWS and DuckLabs were already working together before the acquisition. Since early 2025 the two have collaborated on integrating DuckDB with Amazon S3 Tables and SageMaker Lakehouse, AWS's tools for querying data stored directly in S3. Amazon vice president and distinguished engineer Andy Warfield framed the acquisition around DuckDB's decade-long push to let people, in his words, "work with data with confidence." The strategic goal, per AWS's own announcement, is to make its native analytics tools faster and cheaper, and to turn S3 from a plain storage layer into something customers can query directly.
Part of a bigger pattern
GeekWire's coverage places the deal inside a broader trend of large cloud providers acquiring the companies behind popular open source data tools, rather than building competing products from scratch. Buying the team and the commercial layer around a widely used open source project, while formally leaving governance independent, has become a recurring move for cloud vendors trying to shape which tools become the default for a given workload without alienating the open source community that made the tool popular in the first place.
The founders stay, for now
Keeping Muhleisen and Raasveldt in charge of DuckDB's technical direction, rather than folding the project into an existing AWS team, is a deliberate signal to the open source community that built DuckDB's adoption in the first place. Acquisitions of open source companies often struggle here: once the founders leave or lose authority over the roadmap, contributors and users tend to notice quickly, and forks or competing projects can gain ground. AWS's announcement goes out of its way to name both founders and their continuing role, which suggests the company is aware that DuckDB's value depends as much on community trust as on the code itself.
What it means for teams using DuckDB
For most current DuckDB users, the practical short-term effect should be limited: the license, governance structure, and core team leadership are all staying put. The bigger question is medium-term direction. With DuckLabs' engineers now inside AWS and DuckDB increasingly wired into S3 Tables and SageMaker Lakehouse, expect AWS's own cloud services to get first-class DuckDB integration ahead of, or in addition to, community-driven tooling. Teams building on DuckDB outside the AWS ecosystem should watch whether future feature development starts to favor AWS-specific integrations, even as the core engine itself stays open and free.
