Today, startup analytics platform MotherDuck revealed it has received $47.5 million in funding. That figure includes $35 million in Series A funding and $12.5 million in seed funding, for a total valuation of $175 million.

The company’s founders have worked at some of the most notable organizations in the tech industry, including BigQuery, Amazon Web Services, Databricks and Snowflake. MotherDuck’s solution pairs the functionality of DuckDB with a scale-up approach to collaborative cloud analytics.

Its value proposition, of painlessly querying data where it is, is based on the notion that the size of data involved in such computations isn’t as important as the ease of use. According to MotherDuck CEO Jordan Tigani, who was one of BigQuery’s founding engineers, “Everybody’s focused on this scale-out big data. That’s actually not what’s important. It’s like, can you make it easy? Can you make it simple so people can ask questions and get their answers?”

Several aspects of MotherDuck are designed to do just that. The platform couples serverless computing architecture with what Tigani termed “a hybrid engine that combines local and server side execution.” Consequently, users can query data where it is—at the edge, on laptops, in data lakes or in popular frameworks like Python and Tableau.

Moreover, they can do so without elaborate (and costly) data pipelines, while minimizing data movement and utilizing query optimizations to maximize the user experience.  

MotherDuck: Queries on demand

DuckDB’s lightweight imprint and rapid processing is central to MotherDuck’s query-in-place approach. When users issue queries, MotherDuck spins up a DuckDB instance “in 30 milliseconds or less,” Tigani estimated. “Human reaction time won’t even know it happened.” The instance remains as long as the session lasts, enabling the solution to utilize caching and in-memory approaches to optimize performance. The platform enables organizations to scale up or down as needed. Users can store data in the cloud with MotherDuck “just like you can with Snowflake, BigQuery and Redshift,” Tigani said. “We do have separation of storage and compute so you can scale up the compute independently of the storage infrastructure.”

A key point of distinction between MotherDuck’s method and that of the established cloud data warehouse Tigani named is that with the former, individual users get respective instances. “So, you only have to scale up to the size of each individual user’s workload,” Tigani commented. This approach assists with concurrency and cost reductions, since organizations don’t have monthly costs regardless of how much (or little) they use their data in other cloud warehouses.

Local querying

Many of MotherDuck’s ease-of-use benefits pertain to its local querying capabilities—which are again traced to DuckDB’s utility. Every MotherDuck client will run DuckDB. It’s possible to run the database in browsers and connectors. This functionality enables business intelligence and data science platforms like Tableau and Python to run DuckDB locally, as well. It also means edge devices, such as laptops, can use MotherDuck to run DuckDB locally. “We can actually push data down to the end user,” Tigani revealed. “With DuckDB you can have all of the code locally so you can do a lot of work without hitting any kind of server.”

The possibilities this model enables are significant, if not novel. Users could download a Salesforce report on their laptops as a CSV file, for example, and start asking questions—without configuring data pipelines and moving data to a warehouse. Similarly, developers can ask questions of data locally to begin using it to impact code creation for applications—simply, quickly, and without the complicated infrastructure of pipelines and cloud warehouses. “That’s a lot of work to get some basic questions answered,” Tigani remarked about the pipeline approach.

Ensuing implications

Granted, numerous aspects of data governance, regulatory compliance, and auditing must be answered to fully exploit MotherDuck’s edge computing potential. But its basic value proposition to the enterprise is difficult to deny. Laptops today are more powerful than they ever were before (Tigani called them “supercomputers”). The sizes of data individual users require isn’t always at the scale of big data.

With conventional methods, “If I run a query against Redshift and it takes five seconds, my supercomputer is sitting idle, while this expensive cloud hardware is being used,” Tigani said. MotherDuck is attempting to invert this paradigm to lower costs and make the query process—and data management itself—simple via local computations and minimal data movement.

As the company’s funding announcement and valuation seem to imply, there is an audience with more than an academic interest in the matter.