Stripe, which has emerged as the leading payment infrastructure for the internet, is currently valued at $95B. It's a massive online payment company that you probably have never heard about. It is based on an idea of taking something extremely complex - infrastructure for online payments - and reducing it to 7 lines of code that could be integrated into any online platform and would take care of the entire payment infrastructure. It powers almost all major platforms such as Uber, etc. to tons of small online retailer platforms.
Similarly, we started Netra with a simple idea, understanding the content of every video and making it very simple to leverage the value of video. Videos are rich media, there is a lot of information packed in a few seconds of video content. And understanding the content of video goes well beyond understanding a few objects appearing on screen. It's about getting the context in which these objects appear. It's not just about understanding that there is an automobile present in the video. It's also recognizing if the car is driving through the streets of a crowded city or if the car is driving in the countryside, if it gets met with an accident, etc. Or if it is a street view camera, not just about understanding what objects appeared in the camera, it is also important to narrow down to the few moments when something unusual (an anomaly) happens, like a disruption, a traffic accident, etc.
The most important of it all, object permanence. How long did the car appear in the video to determine the relative importance of the car in setting the context. To solve these challenges, Netra came up with a simple narrative to describe the video <subject> <object> <event> at <location>. Thus how important is a person drinking coffee in a restaurant to the whole video, creates an understanding of the video. But in addition, also provide information such as emotions, or overall description of the video (pets, sports, etc.), or perhaps how similar the two videos are.
Most importantly, we built the analysis as an API - a mere 13 lines of code (a little more than 7, but still good) that our partners can integrate within their platforms to start leveraging the value of video. The use cases can vary - from getting the context of the video to flagging any unsafe parts of the video, getting access to products that appear on-screen, or even tracking the pixels of specific objects. With physical camera feeds the API provides information about any anomalies such as disruptions. All this to enable the video economy. And to do all of this at scale, in real-time for feeds to keep up with the amount of content created in the digital and physical world. This has also enabled our partners to focus on providing additional features within their products and overall, as a result, we see anywhere from 30-50% increase in revenue with an easy step of integrating our API.
Netra aims to enable our partners to leverage the value of the video, in a simple, flexible, and inexpensive way. Like Stripe, Netra has emerged as the platform that few have heard of but has integrated into the video economy, into platforms that are utilizing the value from videos at scale and in real-time.