Where it's based: New York
When it was founded: 2013
Why it's Fierce: Dextro was founded to take advantage of the new field of "computer vision," wherein data on digital content like photos is gathered and analyzed. But the company's technology could push it into a new realm of profitability because of Dextro's ability to make sense of one of the hottest new consumer technologies out there: live streaming, Twitter-friendly apps like the ones fielded by Meerkat and Periscope.
Right now, both apps are pretty much stream-as-you-can, with videos ranked by basic factors like number of active viewers or number of "likes."
"So far, in terms of being able to, say, watch concert streams, amateur musicians, behind the scenes in newsrooms, you can't do that on platform because there's so much new content with minimal tags," Founder and CEO David Luan said in a recent interview with FierceOnlineVideo.
Dextro's new app, Stream, currently ties into Periscope's app. Stream automatically curates the active live streams on Periscope depending on a number of factors sussed out by its computer vision platform. Thus, users can find streams of their favorite musician, regardless of where that stream is coming from.
Luan and co-founder Sanchit Arora initially studied computer vision for robotics applications, according to an Xconomy story. But after seeing a trend in which cameras were being built into robots as well as other types of connected devices, they decided to create a simple API that would let connected devices do "smart things" with the objects detected through a camera. Interest from other developers and companies in their API, which unlike many computer vision platforms can glean data from video, led to the formation of Dextro in 2013.
What's next: Dextro continues to push forward with Stream as part of its overall computer vision portfolio. The company received $1.7 million in seed funding a year ago from 10 investors including Skype cofounder Jaan Tallinn. The company hopes to add Meerkat compatibility to Stream, but that process has not yet started.
"For us, we're trying to understand the universe of say, what all cars look like--putting structure to all this user-generated video content," Luan said.