Intel Corp (NASDAQ:INTC) continues its acquisition spree to build AI and machine learning operations. The latest firm it has acquired is Cnvrg.io which runs a platform for data scientists to develop and test machine learning models. The platform can also be used in tracking and training multiple models as well as run comparisons on them.
Cnrvg to offer a platform for running AI models
Established in 2016 by Yochay Ettun and Leah Forkosh Kolben, the startup is valued at around $17 million following its recent funding round in which it raised $8 million. A statement for Intel indicated that Cnrvg would be an independent company which will continue serving its existing and future clients after completion of the transaction. Some of the company’s clients include ST Unitas, Lightricks and Playtika. The company hasn’t provided the financial details of the acquisition or who is likely to join Intel from Cnvrg.
The acquisition comes a week after the company announced the acquisition of San Francisco-based SigOpt to boost its AI operations. Intel will leverage SigOpt’s technologies in accelerating, amplifying and scaling AI software tools. The combination of the companies’ software technology will offer a competitive advantage by creating value for developers and data scientists. SigOpt also had a platform for running machine learning simulations and models.
Intel sets its sights in next-generation chips
Intel has created an extensive footprint in the AI research and development area. The company has been refocusing its operations around next-generation chips to help it competed with other chipmakers. The focus is around its Mobileye autonomous car business that it acquired in 2017 for $15 billion and acquisition of AI chip producer Habana. The Cnvrg.io platform will offer on-premise, cloud and hybrid solutions offered in free and paid tiers.
The startup competes with the likes of Sagemaker, Dataiku, and Databrikcs and smaller operation such as H2O.ai which are built on open source frameworks. Its premise offers a user-friendly platform where data scientists can concentrate on devising algorithms and testing how they work and not creating or maintaining a platform they operate on.