Kanga – Creating the Ultimate Destination for Gaming Fans
Kanga is an early stage New York-based company creating the ultimate destination for gaming fans. The company’s consumer products help fans discover content through aggregation and real-time metadata and drive user engagement through gamification.
Kanga is an early stage New York-based company creating the ultimate destination for gaming fans. The company’s consumer products help gaming fans discover content through aggregation and real-time metadata and drive user engagement through gamification. Kanga works with esports teams, creators, and publishers and is backed by industry insiders and venture capitalists.
Kanga uses a vast amount of data to provide gaming fans a great content discovery platform. Kanga uses computer vision algorithms to detect data and objects on live streamed video gameplay at scale. These algorithms allow the company to create metadata on the state of games being played in the video on platforms like Twitch, Mixer, and YouTube Live. Kanga use the metadata for two main purposes: 1) to classify streams (e.g. is the game exciting?) to improve discoverability of high quality live content and 2) to create engagement on the stream (e.g. prompt users to predict what will happen next!) in real-time.
I believe that Kanga’s technology will allow the company to capitalize on an opportunity to mitigate the fragmentation within the online streamed gaming content market. This trend of increased fragmentation across gaming distribution platforms is likely to continue to get worse given the proliferation of new content as well as new platforms. Recent examples include Ninja and Shroud, the #1 and #3 streamers on Twitch respectively, moving to Mixer, Discord and Reddit both launching streaming platforms, and TikTok and Instagram becoming homes for gaming highlights and clips. Given this dynamic, I believe that Kanga’s ability to use the game metadata to improve the discovery of content to help users figure out which live content to watch compared to prior methods which only let users follower their favorite gamers.
The computer vision algorithms to detect data and objects on live streamed video game play is among the company’s core assets. The investment necessary in order to build and maintain the app/algorithms is similar to that of other deep technology gaming software development companies.
Website: https://www.kanga.com/
I think this is a fascinating application of data analytics that is at the frontier of our technical capabilities. There is active work in building AI that can analyze and compete in real-time strategic software games (such as league of legends or DOTA). Leveraging this work and ongoing research on computer vision is a very clever way to process unstructured data such as videos or streams.
It’s interesting how kanga is using their algorithms for discovery. I think there will be ample opportunity in the future as the volume of data explodes. Curating the content that is produced and providing “highlights” seems to be a valuable service for customers who are facing an increasingly large choice set of how they spend their leisure time on the internet. For example, streamers will often do multi-hour/day events – having a service that identifies the key moments and creates short compilations would be of value for both content producers and consumers.