Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Amazon completed its acquisition of DuckLabs on 1 September 2026, bringing the team behind DuckDB into Amazon Web Services (AWS). DuckDB itself remains free and open source under the MIT license, with the independent DuckDB Foundation continuing its stewardship.
The immediate product change is in Amazon Aurora PostgreSQL: AWS now embeds DuckDB so Aurora can query live PostgreSQL data—including uncommitted writes—alongside Apache Iceberg and Parquet data in Amazon S3, S3 Tables or AWS Glue Data Catalog through one SQL workflow. The capability is available on Aurora PostgreSQL 17.11 and later and 18.6 and later.
What Amazon bought and when it took effect
DuckLabs announced on 26 August 2026 that it would join AWS. Its follow-up announcement said the transaction process concluded on 31 August and became effective on 1 September 2026. AWS described the deal as the acquisition of the Amsterdam company behind DuckDB.
DuckLabs says its team remains in Amsterdam and continues developing DuckDB, DuckLake, Quack and the broader ecosystem. Founders Hannes Mühleisen and Mark Raasveldt continue leading the team and the open-source project’s technical direction inside AWS. DuckLabs reported more than 30 employees and more than three million DuckDB downloads per day in 2026.
#1 Best Overall
What does not change for DuckDB users
Licensing remains MIT
DuckDB, DuckLake, Quack and related components remain free and open source under their existing MIT licenses. The acquisition does not convert DuckDB into a proprietary AWS-only component.
Independent project stewardship continues
The independent nonprofit DuckDB Foundation continues to steward the project. The DuckDB project announcement said there would be no changes to the roadmap, licensing or governance, while the founders said “the foundations of the project will remain firmly in place.” AWS now employs the core team, but that employment change is separate from the Foundation’s role and the projects’ licenses.
Rank #2
How DuckDB is embedded in Aurora PostgreSQL
In an AWS announcement dated 30 September 2026, AWS said DuckDB is embedded directly within Aurora PostgreSQL. Aurora can therefore evaluate a query that combines transactional tables with data-lake sources instead of exporting the operational data to a separate analytics system first.
Sources that can be combined
- Live Aurora PostgreSQL tables, including uncommitted writes.
- Apache Iceberg tables.
- Parquet files stored in Amazon S3.
- Data in S3 Tables.
- Cataloged data exposed through AWS Glue Data Catalog.
The processing stays within Aurora, according to AWS. That design removes the additional network hops associated with sending data to a separate query engine and avoids creating duplicate analytical copies through an ETL pipeline.
What one-query analysis enables
A real-time dashboard can enrich current orders or account activity with historical lake data without waiting for a scheduled warehouse load. A transaction-oriented application can compare a newly written record with years of archived events. An AI agent can use a familiar PostgreSQL interface to reason over operational and historical data together. These are architectural capabilities; actual latency, concurrency and cost will still depend on the Aurora instance, data volumes, file layout and query plan.
Supported Aurora versions and required setup
| Aurora PostgreSQL release | Minimum version with embedded DuckDB |
|---|---|
| PostgreSQL 17-compatible Aurora | 17.11 |
| PostgreSQL 18-compatible Aurora | 18.6 |
The feature is not described as available on every Aurora PostgreSQL version. Check the engine version before planning a migration or changing a production cluster.
- Run a supported engine version. Upgrade to Aurora PostgreSQL 17.11 or later, or 18.6 or later, subject to your normal maintenance and compatibility testing.
- Provide an IAM role. Configure a role with the
AuroraAnalyticsfeature so Aurora can access the analytics capability and the relevant lake or catalog resources. - Enable the extension. In the target database, create the
aurora_analyticsextension using the privileges required by your Aurora environment:CREATE EXTENSION aurora_analytics; - Expose the lake data. Make the Iceberg tables, Parquet objects, S3 Tables or Glue Data Catalog entries available to the Aurora analytics integration, with IAM permissions covering the specific locations.
- Test representative joins. Validate permissions, transaction behavior and query plans with the same mix of current operational data and historical files that production workloads will use.
AWS has not stated in the announcement that older Aurora releases receive the capability through an extension alone; the engine-version requirement remains a prerequisite.
Architecture before and after the Aurora integration
| Question | Separate pre-integration design | Aurora with embedded DuckDB |
|---|---|---|
| ETL or reverse ETL | Usually requires a separate export, replication or transformation path when the operational database and lake are in different engines; exact requirements depend on the existing design. | AWS positions the feature as avoiding extra ETL pipelines that duplicate operational data. |
| One SQL statement across systems | Normally requires federation, a copied dataset or multiple query steps. | Operational tables and supported lake sources can be combined through Aurora’s PostgreSQL interface. |
| Data coverage | Determined by the separate analytics engine and its connectors. | Iceberg, Parquet in S3, S3 Tables and Glue Data Catalog are listed by AWS. |
| Network movement | Data commonly crosses an engine boundary or is copied before analysis. | Processing remains within Aurora, reducing the extra network hops described by AWS. |
| Data duplication | Analytical copies may be needed to keep a separate engine current. | Queries can read the operational data and lake data in place rather than creating another copy for each analysis path. |
| Supported Aurora versions | Not applicable to this embedded feature. | 17.11+ and 18.6+. |
| Open-source governance | DuckDB is used as an independent open-source project. | The acquisition changes DuckLabs’ ownership, but DuckDB’s MIT license, Foundation stewardship and stated governance remain in place. |
What the acquisition means for different users
Existing DuckDB developers and users
Your licensing terms and the project’s stated governance remain unchanged. The Amsterdam team continues to lead technical development, now with AWS resources behind it. You can continue embedding DuckDB, distributing MIT-licensed code and participating in the existing ecosystem under the same license terms.
Aurora operators
The main decision is whether an engine upgrade and IAM integration justify consolidating operational and lake queries in Aurora. Review workload isolation, permissions, instance sizing and failure handling before moving a dashboard or application path. Embedded execution removes a data-transfer step, but it does not make large analytical scans free or eliminate the need to design efficient Iceberg and Parquet layouts.
Data-platform teams
The capability is most useful where freshness matters and where maintaining a second synchronized copy has been a bottleneck. It is less compelling if your organization already has a deliberately separated warehouse architecture, strict workload isolation requirements or lake formats outside the listed sources.
Bottom line for 2026
Amazon’s purchase gives AWS direct access to the team and technology behind DuckDB, while the project remains MIT-licensed and independently stewarded. The concrete user-facing change is Aurora PostgreSQL’s embedded DuckDB integration: on supported engine versions, one Aurora query can combine current transactional data with Iceberg and Parquet data in AWS storage without first building a duplicate ETL path.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
