AWS acquires DuckLabs: DuckDB, the most beloved in-process analytical database by developers, joins Amazon

Amazon Web Services has just announced the acquisition of DuckLabs, the team behind DuckDB—the open-source in-process OLAP database that has established itself in just a few years as the go-to analytical tool in the data ecosystem. This acquisition raises questions about the future of the project's open-source model.

DuckDB: A Reminder for Those New to It

DuckDB (duckdb.org) is an analytical database (OLAP - Online Analytical Processing) that runs in-process: no server to launch, no daemon to manage. You embed it directly in your Python, R, Node.js, or Java application, or use it via the CLI.

import duckdb

# Analyze multi-GB CSV files in seconds
result = duckdb.sql("SELECT category, SUM(revenue) FROM 'data.csv' GROUP BY 1")

Typical use cases include analyzing large data files (CSV, Parquet, JSON) locally without setting up a Spark cluster or a Redshift instance. DuckDB excels with columnar data (Parquet, Arrow formats) and can process multi-gigabyte files in seconds on a standard laptop.

Its rapid growth in the data science and data engineering ecosystem has been impressive: in just 2-3 years, it has become an almost indispensable tool for Python data analysts.

What AWS Acquired

DuckLabs is the commercial entity built around the DuckDB project (originally a research project at the Centrum Wiskunde & Informatica in the Netherlands). DuckLabs employed the project’s core maintainers and monetized it through MotherDuck, a managed cloud service for DuckDB.

AWS has now acquired both the team (the minds behind the engine) and, implicitly, strategic control of the project—even though DuckDB remains technically under the MIT license, a permissive open-source license.

The Immediate Question

DuckDB is MIT-licensed: anyone can fork it, redistribute it, or use it commercially. In theory, AWS does not “control” DuckDB more after the acquisition than before.

In practice, however, the core maintainers will now work for AWS. Roadmap decisions, feature development, and integration with the AWS ecosystem (S3, Athena, Glue, Redshift) will likely be shaped by Amazon’s interests. This isn’t necessarily negative—AWS’s resources could accelerate development—but it shifts the project’s governance balance.

The precedent set by Elasticsearch (acquired by Elastic, then forked by AWS into OpenSearch when Elastic changed its license) shows how these dynamics can evolve—both ways.

What This Means for You

If you use DuckDB locally: nothing changes in the short term. The code remains MIT, existing versions stay available, and the CLI and Python/R/etc. bindings will continue to work.

If you use MotherDuck (the managed DuckDB cloud): expect a gradual migration to AWS infrastructure and potentially deeper integration with AWS services. Pricing may evolve.

If you’re evaluating DuckDB for a new project: it’s still an excellent choice for local analytics. But if your organization is cautious about AWS dependencies, now is a good time to assess alternatives: Polars (Python, very fast), ClickHouse (server-based but highly performant), or Apache DataFusion (Rust, in-process).

Key Takeaway: DuckDB remains MIT open source—the DuckLabs acquisition by AWS doesn’t change that legally. But the project’s de facto governance is now under Amazon’s influence. Worth monitoring over time, especially regarding roadmap decisions and licensing policy.