Our Series A in Visor: Applying AI to Gaming

In an era where “consumer investing is dead”, there remains one evergreen category with constant innovation, widespread consumer adoption, and deep engagement—gaming. This phenomenon has been supercharged in recent years by the rise of the streaming platforms such as Twitch and the subsequent advent of esports, which has injected new energy and momentum into the ecosystem. Accel has a proud tradition of investing in gaming across multiple fund cycles and partners, leading or participating in investments in Discord, Rovio, Supercell, Razer, and more, so it’s not surprising that we’ve decided to dig in and understand what new market opportunities are emerging.

Gaming has never been more pervasive in our collective cultural consciousness. Twitch has more monthly actives than Netflix or Spotify has subscribers, Antoine Griezmann celebrated a goal in the World Cup finals by doing a Fortnite dance, the top professional esports player makes more money than the quarterback for the Dallas Cowboys, and more than 50 US universities offer varsity esports programs and scholarships.

Today, over 200m+ gamers play AAA game titles such as Fortnite, LoL, DotA, CS:GO, and PUBG every month for an average of 40 hours a month, and over 700m+ consumers worldwide consume gaming video content on a monthly basis. This shift has coincided with the rise of a new emergent genre of games and a fundamental change in how we play them. Instead of linear, single-player games with simple, static objectives, games today are team-based, online, and multiplayer, with competitive play as the focal point of the experience.

Which leads me to Ivan and Anhang, the founders of Visor.gg. It quickly became obvious in our first meeting they had a clear and differentiated insight into this market, both from their own extensive experience as gamers but also from hundreds of customer interviews they had conducted while iterating on product. They saw the growing market opportunity in helping everyday gamers sharpen their skills to compete effectively, but understood that traditional online guides, walkthroughs, or stats websites were obsolete in today’s world of constantly-evolving, dynamic gameplay mechanics and combinatorially large matchup possibilities. Gamers need something better.

Historically, building real-time applications on top of gaming experiences has been prohibitively difficult; APIs for even the most popular AAA games are asynchronous, incomplete, rigid, and differ from game to game—if an API exists at all. To solve this, Visor has built the first real-time, data pipeline for gaming, using custom computer vision and machine learning models which are both scalable and low-latency. I can’t overstate the difficulty of this technical accomplishment, as well as our anticipation at what applications Visor and other developers will build on top of it. Our first application, launching today for Overwatch, provides real-time, personalized, contextual insights and longitudinal analysis for gamers as they’re playing.

The team behind Visor is special and one of the scrappiest early stage teams I’ve seen. Ivan and Anhang met while attending UC Berkeley (Go bears!). Ivan previously spent time investing at a hedge fund while Anhang led engineering teams at Facebook. Perhaps more importantly, they are both serious gamers. Ivan was previously ranked in the top 1% of HoN players, and Anhang played competitive DotA from 2007-2010. Both are now avid Overwatch and Fortnite players. The rest of the early team is just as exceptional, with talented engineers and designers from Airbnb, LinkedIn, Facebook, AppDynamics, Nextdoor and Humble Bundle who are equally passionate about the problem space.

Visor represents an exciting new chapter in gaming—one that opens the door for more real-time in-game experiences and a more efficient flow of information through the ecosystem. It unlocks and democratizes innovation on top of gaming, which has typically been reserved for a few publishers with distribution. We’re excited to see what you build.