So here's how it works.
Here's me telling the local TV channel about the project.
I started work on this project on 1st Jan 2016, because I thought there should be a more neutral way of news storytelling, and what's more neutral than an AI. I worked on this project for 33 weeks, grew the team to 15 people (both remote and full time) but later had to shut down the project of copyright issues.
I made this project at an amazing hackathon, hosted by facebook Seattle. Good food, great people and amazing ambience. Best hackathon I've ever attended.
It's a program that uses OpenCV facial recognition to identify repetitive faces in a video and compress the video by cropping the relevant portions of the video, here's a slideshow for more info.
In May 2016 this project got picked the amazing Amazon "catalyst" program and now we're working with amazon to make educational videos accessible to all bandwidths of internet.