Amazon Acquires DuckLabs, Pledges to Keep DuckDB Open Source for AI Agents

Amazon Announces Acquisition of DuckLabs. Headquartered in Amsterdam, DuckLabs is the company behind the analytical database DuckDB.

According to reports, Hannes Mühleisen and Mark Raasveldt, creators of DuckDB and co-founders of DuckLabs, will join Amazon Web Services (AWS) to continue leading the team and overseeing the technical direction of the open-source project.

In an announcement on the DuckLabs blog, Mühleisen and Raasveldt explained their reasons for joining AWS:

Joining AWS gives us the resources and reach to bring this technology to more developers and organizations, and to explore ideas at a scale that would have been difficult to achieve independently.

All parties involved in the deal assured the community that the future of projects maintained by DuckLabs remains unaffected. DuckDB and all components of its "Duck Stack," such as DuckLake and Quack, will remain free and open source under the MIT license. The non-profit DuckDB Foundation will also continue to oversee these projects.

Mai-Lan Tomsen Bukovec, Vice President of Technology at AWS, explained the problem DuckDB was originally designed to solve in the announcement:

DuckDB aims to deliver extremely fast performance for the over 90% of data queries worldwide that typically involve processing 1TB or less of data in analytics and dashboard scenarios.

She believes that DuckDB’s optimizations and performance are particularly important for agentic AI workloads. AI agents often inspect data and run small exploratory queries before deciding on their next action. This pattern makes DuckDB naturally suited for AI use cases.

The community has generally reacted positively to the news, though some concerns remain. On Reddit, adappergentlefolk wrote:

I hope this doesn’t slow down their pace of innovation, but I’m glad they found a way to get rewarded for their work while keeping most of the projects open source under the foundation.

Andy Warfield, Distinguished Engineer at AWS, further detailed how the relationship between AWS and DuckLabs matured prior to the acquisition in a post on Werner Vogels’ blog. He noted that when the two teams first connected, the S3 team was developing S3 Tables. Since then, collaboration has deepened. He explained:

Over the past two years or so, our teams have worked increasingly closely together, expanding the scope of our collaboration. We explored using AWS Lambda as a matching primitive for quickly launching DuckDB queries, and improved integration between DuckDB and AWS databases and analytics engines through DuckDB’s evolving ATTACH and CONNECT commands.

Warfield views the deal as an opportunity to expand the impact of the two co-founders rather than absorb their work:

Hannes and Mark’s vision aligns closely with how Amazon thinks about data, so we want to provide them with the broadest possible opportunities to not only steer DuckDB but also influence how cloud analytics are built and delivered. Therefore, after many discussions... we jointly decided to bring the DuckLabs team into AWS.

Following the acquisition, Amazon Aurora PostgreSQL became the first service to launch user-facing features powered by DuckDB. Esra Kayabali, Principal Solutions Architect at AWS, explained the type of workload this new feature aims to simplify:

Previously, if an application needed to combine recent transactional data from Aurora with historical data stored in Amazon S3, a common approach was to build reverse ETL pipelines. This resulted in data duplication, increased infrastructure costs, and required ongoing engineering effort to keep all data synchronized.

With DuckDB directly integrated into Aurora PostgreSQL, users can now combine real-time operational data with historical data from Apache Iceberg tables managed by the AWS Glue Data Catalog, as well as Parquet and Iceberg data stored in Amazon S3 and S3 Tables, using a single query. These queries are processed entirely within Amazon Aurora, eliminating the need for additional network hops or ETL pipelines, thereby minimizing configuration and resource requirements.

This feature is available in all AWS commercial regions and AWS GovCloud (US) at no additional cost. Customers pay only for the incremental Aurora compute resources and S3 requests generated by the queries.

Original article:

https://www.infoq.com/news/2026/10/amazon-acquires-ducklabs/