 |
VOOZH | about |
|
VOOZH | about |
We’re so glad you’re here. You can expect all the best TNS content to arrive Monday through Friday to keep you on top of the news and at the top of your game.
Check your inbox for a confirmation email where you can adjust your preferences and even join additional groups.
Follow TNS on your favorite social media networks.
Become a TNS follower on LinkedIn.
Check out the latest featured and trending stories while you wait for your first TNS newsletter.
During the recent MCP Dev Summit North America held in New York City, The New Stack sat down with Till Döhmen, MotherDuck’s AI lead, to discuss how MotherDuck started its MCP journey and how even non-technical workers at the startup can now use agents to interact with their own data rather than relying on dashboards.
MotherDuck, a startup backed notably by Tomasz Tunguz and able to hire the inimitable Carly Spoljaric, is commercializing the popular open-source DuckDB analytical database.
But more interesting were Döhmen’s comments on how MotherDuck liaises with the DuckDB foundation, which warehouses — pun intended — DuckDB IP and leads the open-source project itself.
“We have a very close collaboration with DuckDB Labs … specific things that should be added to DuckDB.”
“We have a very close collaboration with DuckDB Labs,” Döhmen said, explaining that if there are “specific things that should be added to DuckDB,” MotherDuck can go and talk to the open-source maintainers about expanding the core project.
Why is that useful? Because, as Döhmen explained, his startup operates “the largest fleet of DuckDB databases in the world,” allowing MotherDuck to stretch the limits of what the database can handle. That information is gold to the folks keeping DuckDB up to date in our era of ever-larger datasets and heavier, more frequent analytical runs.
Blast from the past: TNS covered MotherDuck in 2023, noting at the time that the startup had “undergone three rounds of funding in approximately 15 months,” raising around $100 million in known capital in the process.
Enter commercial tension: TNS wanted to know what was stopping the startup from improving DuckDB on its own, rather than federating information back to the foundation in charge of the core project — why not hoard that information and potentially offer a more differentiated product?
Rude, yes, but I was curious. Döhmen was emphatic that his company was not interested in forking DuckDB. Instead, he explained that MotheDuck takes advantage of DuckDB’s extensibility to amend it as needed without issue (specifically citing its query planner).
“But if we encounter things that happen that are inductively core that we would like to change,” it’s a different story, he said.
It’s not rare to see companies built atop open-source projects. Databricks was founded by the team behind Apache Spark. Comfy was born out of an open-source project, and recently raised a $30 million round. The list goes on.
But those successes politely obfuscate the back-and-forth between commercial arms that take an OSS project to market and the folks doing the messy work of maintenance and improvement.
“It’s a partnership, a dance between two groups with very different bosses.”
It’s a partnership, a dance between two groups with very different bosses. Still, it seems that, as with Databricks, MotherDuck is able to both drive market adoption of open-source software and achieve venture-attractive growth.
