Revery.ai recently joined TSVC’s portfolio, enriching the firm’s vision in deep-tech and computer vision. Revery.ai was founded in October 2020, by three talented computer science PhD students from UIUC.
Three co-founders Kedan Li, Jeffrey Zhang and Min Jin Chong. Image Credits: Revery.ai
The company targets the clothing fitting experience in the e-commerce fashion industry, where try-on and visualization are difficult to manage remotely. The current solution relies on providing a few carefully modeled photos; however, the large psychological gap between expectation and what customers actually receive can lead to high return rates. Revery.ai creates a scalable way to visualize outfits on people through the latest deep learning and computer vision algorithms. Their product can customize the fitting experience and a vendor can apply it to millions of products with less than a week's effort.
Virtual try-on. Image Credits: Revery.ai
This is not a crowded space, though there are competitors, such as DRESSX and Zeekit. DRESSX targets a slightly different problem: rather than consumers' virtual try-on experience for physical purchases, they appear to focus on fashionable but digital clothes for virtual avatars. They state that they intend to invite and/or sign digital designers for collaborations and create a marketplace, akin to a “digitalized” Farfetch. Zeekit was acquired by Walmart this year. Although deal terms were not disclosed, they had raised a total of over $16 million since 2014 according to TechCrunch. Both Zeekit and Revery.ai focus on the virtual try-on experience in the retail sector.
TSVC gets involved with founders very early in their companies' life cycles, when aspects of their business model are often still in flux. We share our perspective and experience to add to their own thinking, provide another perspective on how they can grow. While we were impressed with Revery's sophisticated solution, Dr. Chun Xia, the TSVC partner leading this deal, also felt that the company may face a growth ceiling and intense competition if they limit themselves to the customer-facing virtual fitting space.
While virtual cloth fitting is an immediate entry point for Revery.ai, what Dr. Xia believes in is the enormous potential of their technologies applied to and transforming the social network in the long term. To make this point of view more specific, IMVU is a great example. IMVU, founded in 2004, is a social platform in the metaverse space. Users from IMVU can customize and create their own 3D avatars, meet, chat, and play games with other people in the virtual world. The company reveals that, despite the pandemic, it saw record growth with over 7 millions active users and 400,000 user generated contents every month.
Dr. Xia has been following the trends of the metaverse industry in the past twenty years and what he sees as the potential of Revery.ai is the virtual cloth fitting technologies applied to the meta-human virtual characters, or virtual idols/KOLs.
What is virtual KOL? Take a look at the below figure. Imma from Japan has over 345k followers on instagram, Bermuda from the U.S. has over 284k, and Ayayi from China - over 112k on Xiaohongshu (a Chinese social platform). Imma, perhaps the most successful and impactful character, has become the brand ambassador of Watsons and KFC China.
In the current practice, these virtual influencers are mostly produced by companies from the digital marketing / advertisement industry. They are professional and resourceful. However, the current practice requires very dedicated efforts and high costs to create and maintain even just one such character. Dr. Xia believes that such industry can be transformed by Revery.ai’s revolutionary and sophisticated technology and make it available to all end users. Every person can create its own meta-human character, explore different fashion styles, virtual possess and dress clothes and accessories. A new lifestyle in the future will be sharing photos and short videos of yourself on social media - “physically” being at any corner of the world and “physically” dressing any style of clothes, without actually going to the location and possessing the items.