CEO Florian Douetteau's leadership has helped make Dataiku a billion-dollar startup with customers ... [+]
At a time when tech giants such as Google and Amazon are weaving artificial intelligence throughout their company fabrics, data remains siloed at many companies outside of tech, left to the work of employees with the word “data” in their job titles. Data science startup Dataiku thinks it has the tools for companies in retail, finance and the like to utilize data like their tech peers. “I suspect in ten years we won’t be using spreadsheets,” says CEO Florian Douetteau.
Instead, he hopes enterprises will turn to Dataiku’s web-based software to take themselves on the path to artificial intelligence. The New York City startup on Monday announced that it had raised $100 million in a Series D round led by Stripes and Tiger Global Management. The company, which was previously valued at $1.4 billion after Alphabet investment arm CapitalG became an investor last December, would not disclose its new valuation, but said that it was “still a unicorn.”
“We believe Dataiku is about the future of AI in the enterprise, which is a future that is going to be more collaborative and more inclusive,” Douetteau says. “Anyone within the enterprise will be able to process data or massage data to get something out of it, whatever their skills.”
Dataiku was founded in 2013 by Douetteau and three fellow Frenchmen: Marc Batty, Clément Stenac and Thomas Cabrol. Rather than providing a specific solution, Dataiku’s platform makes it easier for data scientists and other employees to make use of unstructured data for their own purposes. The software allows workers across different teams to collaborate across the different parts of the process, such as the scrubbing, wrangling or analyzing of data. “We provide tools to enable analysts, data scientists and employees in general in order to enrich data rather than developing a solution by themselves,” Douetteau says. “It might be a function of origin—being French, we are lazy so we want to do as little as possible.”
The startup has cast a wide net in terms of both industries and locations, with more than 300 customers across North America, Europe and Asia including General Electric, Morgan Stanley and Pfizer. Levi’s uses the software’s predictive modeling and machine learning capabilities to create a personalized recommendation system for shoppers, while Mercedes-Benz uses it to forecast business performance.
The company, which features on the latest editions of Forbes’ AI 50 and Cloud 100 lists, faces stern competition in its space. Various startups—including fellow AI 50 honorees DataRobot and Domino Data Lab—and more established companies such as Alteryx and Tibco Software are focused on providing machine learning tools to help companies do more with data. The market space is wide, as any existing company without a dedicated data science setup could be a potential customer. Other businesses have cropped up to provide similar software for specific industry with the thinking that their offerings can be better tailored to the sector-specific needs of their customers.
Dataiku, however, says its diversification has been crucial to its positive trajectory during the Covid-19 pandemic. The financial effects of struggling transportation and manufacturing customers have been more than offset by companies in other sectors that are accelerating their digital transformation timelines. The startup has added more than 100 employees and “dozens” of new clients this year, including oil services company Schlumberger.
The fresh capital—which came with additional participation from past investors CapitalG, Battery Ventures, Dawn Capital, FirstMark Capital and Iconiq Capital—will go towards the continued growth of the company and its offerings, though the company leadership says it is already feeling validated by the original strategy to target a wide range of geographies and use cases. “We don't have to go and acquire other companies to complement parts of the product that are missing,” says chief customer officer Kurt Muehmel.
“For us, it’s not, ‘How do you build the best possible X or Y for a given use case?’” he adds. “It’s how do you solve the meta-problem one level up, which is: How does every company equip itself with the ability not to be reliant on a service provider or a particular vendor, but to build those solutions themselves.”
