DuckDB Now Runs Inside Amazon Aurora PostgreSQL as AWS Ships Its First Native Integration Since Acquiring DuckLabs

AWS has rolled out its first native DuckDB integration since acquiring the database's parent company, embedding DuckDB directly inside Amazon Aurora PostgreSQL. The feature, announced on October 17, lets a single SQL query join live operational tables with historical data stored in Apache Iceberg tables, Parquet files, Amazon S3 Tables and the AWS Glue Data Catalog. It is available in all commercial AWS regions and AWS GovCloud (US) at no additional charge, with customers paying only for incremental Aurora compute and standard S3 requests.

The integration follows Amazon's acquisition of DuckLabs, the Amsterdam-based company behind the open-source analytical database. DuckLabs announced it would join AWS on August 26, the transaction concluded on August 31 and became effective on September 1. Co-founders Hannes Mühleisen and Mark Raasveldt continue to lead the team and the project's technical direction inside AWS, and the company's more than 30 employees remain in Amsterdam. DuckLabs reported more than 3 million DuckDB downloads per day in 2026.

DuckDB, DuckLake, Quack and related projects remain free and open source under their MIT licenses, with the non-profit DuckDB Foundation continuing its stewardship role. "Joining AWS gives us the resources and reach to bring this technology to many more developers and organizations, and to pursue ideas at a scale that would have been difficult for us to reach alone," Mühleisen and Raasveldt wrote in a blog post.

The technical payoff is the removal of reverse-ETL plumbing. Previously, applications that queried live Aurora tables alongside historical records in S3 had to copy data through pipelines that duplicated records, raised infrastructure costs and needed constant maintenance to stay in sync. Now a query can include the operational side's current state — AWS specifically notes that uncommitted writes are visible — so dashboards and analyses see the database as it is being written, not a stale replica.

AWS vice president Mai-Lan Tomsen Bukovec said DuckDB was built for the 90 percent-plus of data queries that run on one terabyte of data or less, and argued those optimizations are particularly relevant for agentic AI workloads, where agents inspect data structures and run small exploratory queries before deciding what to do next. The lightweight, in-process design makes it a natural fit for that pattern, she said.

Adoption requires Aurora PostgreSQL 17.11 or later in the 17 series, or 18.6 or later in the 18 series, an IAM role that includes the AuroraAnalytics feature, and the aurora_analytics extension enabled in the database. Clusters on earlier minor releases are not covered.

The community's reaction has been cautiously positive. One widely upvoted Reddit comment captured the mood: "I hope this doesn't slow the pace of innovation they have been doing, but I'm glad they found a way to both cash in on their work yet keep most of it OSS under the foundation." AWS distinguished engineer Andy Warfield wrote that the two teams collaborated for roughly two years before the deal, beginning with S3 Tables and extending through DuckDB's ATTACH and CONNECT commands.

The deal is a case study in how hyperscalers now absorb open-source infrastructure: acquire the commercial entity, leave the license and foundation untouched, and wire the technology into first-party services where it compounds. For developers, the near-term win is simpler architectures for mixed operational-analytical workloads; the open question, as always, is how much of the roadmap bends toward the new owner's ecosystem over time.