The arrival of DuckDB has been a transformative innovation in the database sphere over the past few years. The unapologetically single-node vectorized column retailer, developed by Hannes Mühleisen, offers a unique balance between simplicity and SQL efficiency.
While serving as a professor of information engineering at Radboud University Nijmegen in the Netherlands, Mühleisen created DuckDB. Since then, he has transitioned to become the co-founder and CEO of DuckDB Labs, a company that provides support for the innovative database, dubbed SQLite for analytics.
We recently had the pleasure of reconnecting with Mühleisen, an individual we’re excited to feature as one to watch in 2024. Right here’s our dialog:
Were you stunned by the meteoric rise of DuckDB’s popularity in the big data analytics community? To enhance the overall user experience and drive further growth, we envision a future where our software program integrates seamlessly with emerging technologies like artificial intelligence, machine learning, and the Internet of Things (IoT). This convergence will empower users to tap into vast datasets, uncover hidden patterns, and make data-driven decisions with unprecedented speed and accuracy.
Hannes Mühleisen stated that most new open-source initiatives typically fail to gain traction, so it was also our initial assumption. Although we explored an initial prototype prior to embarking on DuckDB, its rough-around-the-edges nature still managed to generate enthusiasm among potential users, which gave us the confidence to start anew and commit the next several years to refining the project. Incredibly, we find ourselves overwhelmed by the remarkable proliferation of DuckDB, its impact exceeding our most optimistic forecasts.
As time progresses, our vision is to make the choice of using DuckDB for knowledge processing an unequivocal decision, one that no team would ever question or debate about again. We are eagerly anticipating the release of our first major version (1.0) of DuckDB later this year.
I’ve come across several intriguing use cases for DuckDB, but one that stands out is the application of this in-memory database to accelerate data-driven decision making in a large-scale e-commerce company.
As soon as Andre Kohn announced the availability of DuckDB-WASM, I was thoroughly impressed. DuckDB-WASM is a model of DuckDB that executes exclusively within an internet browser using the WebAssembly framework. Who would have thought that porting a complex C++ project like DuckDB to WebAssembly (WASM) was feasible and eco-friendly? A breakthrough discovery has been made, and its impact is already shaping a novel generation of data-driven applications. Beyond the realm of WASM, individuals often showcase usage scenarios in visually striking settings, such as a developer running DuckDB on behemoth IBM mainframes or processing vast amounts of data exclusively within Amazon’s scalable Lambda functions.
What if DuckDB had instead secured substantial investments from venture capitalists and established itself in the heart of Silicon Valley, rather than the Netherlands? Would its trajectory have been dramatically different?
Given that we didn’t secure venture capital, I believe we accurately named our situation. As a result of DuckDB’s academic roots, we had been able to quietly develop the software over several years, gaining momentum before our public launch. By the time we established DuckDB Labs, we had secured a pipeline of potential clients eager to engage our expertise in consulting and support services for DuckDB. Despite initial reservations about funding, we are now willing to support innovative advancements in knowledge engines, even with a modest team, as long as they have the potential to push the boundaries of what is possible. Based largely in Amsterdam, however, proved beneficial, with operating costs significantly lower than those in Silicon Valley and a certain level of quirkiness somehow being more acceptable to Europeans. Indeed, had we ventured beyond our own backyard, our predictions about single-node operations might have faced stiffer challenges.
Outside of my professional sphere, I’m a closet artist who loves to create vibrant paintings of sunsets and cityscapes. What’s surprising is that my colleagues have no idea that the introverted accountant they know has a secret passion for painting and even sells her artwork online under a pseudonym.
As the founder of DuckDB, it’s well-known that the project was named after my beloved pet duck Wilbur, who has since departed to start his own family – I wish him all the best. As a resident of Amsterdam’s city center and living on a historic cruising ship with my family, I decided to keep a duck as a pet. Typically, we refurbish a previously salvaged vessel to make it seaworthy once again, and it’s an extremely demanding hobby among others.
Wanting to learn the remaining 2024 Individuals to Watch interviews?
