Affine Systems is a startup whose approach to video search isn’t just about the results, it’s also about what’s after that. The company wants you to find what you are looking for, but then bucket those results in a way that will lead you to discover and watch more videos.
There is no shortage of companies offering video search. Like EveryZing and Nexidia, Affine is a B2B play targeting premium content companies looking to provide search services and become a gateway to discovering more video.
Similar to VideoSurf, Affine uses facial and object recognition to find what you are looking for. When you search for George Clooney, you get a number of videos featuring the movie star, but then Affine goes one step further to break down and bucket other results based on what it thinks you might be interested in from that search. So Affine can recognize Clooney and Matt Lauer’s face and know that it’s an interview they did and point you to that. Or it can recognize a CNN logo on-screen for any news clips featuring Clooney on that news network.
The idea is that people show up to a site looking for something specific, but you can use that specific search as a pivot point into something else (a similar principle in mathematics is where the company gets its name). Affine CEO Michael Sullivan believes that this approach will lead to better monetization of content. The company will make money based on the amount of traffic it drives, either through a fixed price per view or an ad split.
Right now, Affine is in beta and hasn’t announced any customers, though Sullivan says those will be coming in the next few months. The company has five employees and recently moved from Cambridge, Mass., to the Menlo Park, Calif. offices of its VC firm Highland Capital Partners. Affine raised an undisclosed Series A in December 2007 and says it has enough money to last until 2010.