Microsoft acquires Bonsai to help with its artificial intelligence push

  • Microsoft's corporate venture arm M12 invested in Bonsai last year.
  • Terms weren't disclosed.
Microsoft CEO Satya Nadella at a company event in Washington in 2014.
Source: Microsoft
Microsoft CEO Satya Nadella at a company event in Washington in 2014.

Microsoft on Wednesday said it has acquired Bonsai, a small artificial intelligence start-up. Terms of the deal weren't disclosed.

Microsoft has increasingly bet on AI and has sought to commercialize ideas its own researchers come up with, a strategy also employed by Amazon, Google and other big technology companies. By buying Bonsai, Microsoft gains a product that has picked up some business use to help with the push.

Bonsai's software draws on a trendy approach called reinforcement learning, which involves training systems to yield better outcomes through trial and error. After experts highlight important areas in the data and train an AI model, it can be incorporated into corporate applications.

The system can be deployed atop Microsoft's Azure public cloud. Microsoft will "make sure it's running great on Azure," Gurdeep Pall, the company's corporate vice president for business AI, told CNBC in an interview.

Bonsai was founded in 2014 and has around 42 employees, with offices in Seattle and Berkeley, California. It had raised $13.6 million in venture funding, according to Crunchbase, from firms such as ABB Technology Ventures, New Enterprise Associates, Samsung, Siemens and Microsoft's own corporate venture arm, which currently goes by the name M12. Bonsai CEO Mark Hammond worked for Microsoft in the late 1990s and early 2000s.

Hammond said Bonsai has engaged with companies dealing in oil, ventilation systems, robotics and cars. In an engagement with Siemens it was able to automate a previously manual process and make it happen 30 times faster, he said.

Microsoft bought another small AI company, Maluuba, in early 2017. A year before that it bought start-up SwiftKey, which has relied on AI to boost the capabilities of mobile virtual keyboard applications.