Daily Management Review

Twitter gets 'Magic Pony'


06/21/2016


Twitter has acquired Magic Pony Technology, a British start-up company specializing in machine learning technologies.



keiyac via twitter
keiyac via twitter
Twitter bought another startup that is engaged in machine learning. In 2014, the company acquired Madbits, in 2015 - Whetlab, and British Magic Pony Technology is the third in a row.

Magic Pony works on machine vision – a technology that enables artificial intelligence to understand what is on the image. Computers can analyze and improve the quality of images and low-resolution video, and automatically create landscape for computer games and virtual reality systems.

Magic Pony works on the following principle: a user uploads two copies of an image in high and low resolution. The artificial intelligence analyzes them and, basing on the received knowledge, "learns" to improve quality of other low-resolution images. Also, the algorithm improves video quality in real time, in particular, helps increase picture sharpness.

Magic Pony Technology Team will join Twitter’s subdivision Cortex, dealing with artificial intelligence.

The company has filed more than 20 patent applications of its technologies. The startup’s employees will join Twitter team. According to TechCrunch, cost of the transaction was $ 150 million.

Twitter had previously stated that one of the project’s priorities is to intensify working with video. Recently, the microblogging network signed a deal with the US National Football League, according to which Twitter may broadcast live games of 2016 season on Thursdays. 

Twitter also owns video streaming service Periscope. Purchase of Magic Pony Technology can help improve quality of smartphone-shot streaming videos hosted on Twitter.

Before Twitter decided to purchase Magic Pony, the startup had received funding in the amount of $ 6.31 million in two rounds from Octopus Ventures and Balderton Capital. In the last month, it held a regular round, during which the company was valued at 21 million pounds.

source: techcrunch.com