Hedge Fund Billionaire Invests in Swedish AI Startup Peltarion
(Bloomberg) -- Euclidean Capital, the family office of hedge fund billionaire James Simons, is leading a $20 million investment in Peltarion AB, a Swedish startup whose software platform helps companies use artificial intelligence.
The new funding values Peltarion at $110 million, Luka Crnkovic-Friis, the company’s co-founder and chief executive officer, said in an interview. Also participating in the round are existing investors EQT Ventures and FAM, a holding company owned by three foundations established by Sweden’s prominent Wallenberg family.
Simons, a mathematician and former code breaker who founded hedge fund Renaissance Capital, has an estimated net worth of $16.5 billion, according to the Bloomberg Billionaires Index. News of the investment was first reported by the Swedish publication DI.
Crnkovic-Friis said that Peltarion, which is headquartered in Stockholm and has offices in London, Oslo, and Barcelona, is planning to open an office on the East Coast of the U.S. in the coming months and will start ramping up sales efforts there.
Peltarion’s software platform helps companies build algorithms using neural networks -- a kind of machine-learning loosely based on the human brain -- and is among a handful of firms looking to make it easier for companies make their own AI, even if they lack the relevant experience or coding skills.
Among its competitors is Petuum, a Pittsburgh, Pennsylvania-based company founded by Carnegie Mellon University computer scientist Eric Xing. Microsoft Corp., in September 2018, acquired California startup Lobe, which built a graphical programming interface for computer-vision tasks.
Facebook Inc., Amazon.com Inc.. Alphabet Inc.’s Google, International Business Machines Corp. and Microsoft have all begun offering pre-trained algorithms and machine learning toolkits. They are doing so, in large measure, in the hope of convincing these businesses to buy their cloud-computing offerings. But Crnkovic-Friis said the tools these companies offer are still too complex for inexperienced users and often not customizable or robust enough to be applied to real business use cases.
"A lot of these things are fine if you want to build a proof of concept, but they won’t work if you need to use them every day for critical business functions," he said.
Peltarion was founded in 2004 as a specialized consultancy that helped the likes of Tesla Inc., General Electric, and the U.S. space agency NASA use neural networks.
In 2016, the company shifted its strategy away from bespoke projects and toward building a much broader software platform aimed at bringing AI to as many businesses as possible. Customers have since used its products to build a number of different AI applications, from systems that help doctors to segment cancer tumors to algorithms that help farmers optimize crop yields.
Crnkovic-Friis said pickup of Peltarion’s software platform since it launched in June 2018 had been "good" but that many businesses were far less ready to use neural networks -- in terms of having their data organized and understanding what they wanted to build -- than he had anticipated.
"More often than not, they need a bit of help," he said. "A deep-learning platform, the gap there, may be too big for them to bridge on their own." Deep-learning is the kind of neural network-based machine learning most commonly used today by companies and involves algorithms that learn from data.
To help win over more customers, he said, Peltarion has turned to partnerships with consulting firms. It has 16 in total, including Sopra Steria and BearingPoint.
©2019 Bloomberg L.P.