AI Becomes Fashionable with these 10 Startups
All we ever learned about fashion came from watching and re-watching Zoolander until we could perfect the Blue Steel look (see below). In other words, at Nanalyze, we’re not exactly a bunch of metrosexual males squeezed into skinny jeans made in Japan for $595 a pop. So it’s not really surprising that a machine knows more than we do about what accessories would work best with a powder blue V-neck sweater. Or even what color powder blue is. What we did find surprising is how many AI fashion startups are in the e-commerce space these days.
Now, applying algorithms to help customize the shopping experience isn’t new. It’s what’s built Amazon into an e-commerce giant, even if its recent streak of profitability has little to do with retail. Amazon uses machine learning for product recommendations. It relies on its vast amounts of data, analyzing specific user behavior as well as purchasing preferences among its millions of consumers, matching similar products to similar customers.
The same approach can certainly work for someone in search for a pair of sneakers. But fashion itself is different; it’s particularly visual. Many of the fashion-forward startups we’ve found are developing more sophisticated algorithms focused on computer vision. Computer vision is a field of AI dedicated to designing machines with the ability to process, understand and use visual data just as humans do.
Below you’ll find companies rolling out entire search platforms, bots and even AI fashion designers. Maybe we won’t have to wait until the 25th century to look as cool as Buck Rogers and Col. Wilma Deering.
10 AI Startups in Fashion
Mad Street Den seems to be generating more buzz than a strain of Kosher Kush. Based in California and India, the startup from husband-and-wife team Anand Chandrasekaran and Ashwini Asokan took in an undisclosed amount of Series A funding last year after a $1.5 million Seed round in 2015. Sequoia Capital, with a penchant for picking winners, led the Series A round. Mad Street Den refers to its AI fashion brand as Vue.ai.
Update 04/24/19: Vue.ai has raised $17 million in Series B funding to equip online retailers with AI smarts. This brings the company’s total funding to $27 million to date.
As Asokan told Fast Company:
It’s basically an end-to-end AI-assisted onsite or on-app set of products. So people are browsing a site or looking up on their app and we’re basically trying to understand what they are looking for, at the starting levels, of color pattern necklines, sleeve length, styles. We’re trying to look at user behavior, history, and their cohort. We’re looking at it at a completely meta level. I think, traditionally, people have looked at it as big data, but we’re actually looking at it and even the visual aspects of every little thing that the user is doing. At the end of the day, fashion is super visceral.
The company claims its clients—including retailers like Levi’s—have reported customers spending three times as long shopping online using its AI fashion platform and doubling the number of items that end up in the shopping cart. The company has expanded its machine-learning algorithms to furniture and is rapidly increasing its staff.
One of the AI 100, as judged by data research firm CB Insights, Mode.ai also emphasizes a visual shopping experience using an AI fashion bot that companies can integrate on their website. No details on its exact funding, but it lists Baseline Ventures, Aspect Ventures, and Asperon as primary investors. The AI fashion bot can search visually and textually from millions of product photos across thousands of e-commerce retailers. It recently launched a direct consumer product through Facebook messenger that you can play with if you’re one of the billion or so people with an account. It didn’t do so well with our Buck and Wilma photo in terms of results, as you can see below:
An app-based product that does a similar visual search using AI fashion algorithms comes from GoFind.ai. No details either on their funding, but the startup came on our radar through AI chip maker Nvidia (NVDA), which is one of the hottest stocks out there. Using Nvidia chips, San Francisco-based GoFind launched last December and reportedly wants to grow into the virtual and augmented reality space.
Founded by a former Saks Fifth Ave executive, Propulse Analytics took in $1.4 million in Seed money in February. Its AI fashion platform asks why pay a high-end sales expert $200K a year when you can automate the process. Its algorithms analyze product images, considering 39,000 variables simultaneously, going beyond the usual purchase and browsing history to learn about a customer’s tastes. One client, Frank + Oak, told TechCrunch that his company is seeing 2.5X conversions with a new recommendations page powered by Propulse.
ViSenze is another company applying AI for visual search platforms in the e-commerce space, including fashion, jewelry and interior design. The Singapore-based startup has raised $14 million, including a $10.5 million Series B last September. Its AI fashion customer list doesn’t include any companies we’re familiar with—but then again, we do most of our shopping at Goodwill. The platform allows customers to search for products using uploaded images, and its algorithms can automatically add relevant tags to products at scale for text-based searches. You can go to its demo site to see how it works. It really nailed our Derek Zoolander image:
The company has also developed an advertising product that drops clickable ads right into video. Love what your favorite soap opera star is wearing? Just click on the pop-up ad and buy it right then and there. You can see that demo here.
Update 2/20/19: ViSenze has raised $20 million in Series C funding to develop its software through partnerships with smartphone makers including Samsung, LG, and Huawei. This brings the company’s total funding to $34.5 million to date.
Another company in Asia, Omnious in South Korea, seems to be peddling a similar AI fashion platform. It recently raised more than $250K. It uses visual search and image-recognition technology to return results from a company’s online catalog, as well as add relevant tags to merchandise, just like ViSenze.
Another Nvidia chip customer, TheTake.ai is also taking a page from ViSenze (or is it vice versa?) by using machine learning and machine vision technology to take product placement in video to a whole new level. The New York-based startup has $2 million in Seed money. Its technology identifies products and locations shown in movies or television shows, then scours the interwebs for similar products for sale. It has a B2B platform and a consumer-facing website where you can get dolled up just like a Kardashian.
The app is also even more useful for trying to avoid looking like a Kardashian.
London-based Thread focuses on AI fashion for men (though recently added women’s fashions), backed by $16.3 million, including a $5.8 million Series B last June. Thread combines HI (human intelligence) with AI. The process starts out with customers completing a short questionnaire that includes choosing images of stylish and fit metrosexual males in various attire that most fits their deepest desires. A human stylist then gets involved, assembling a thumbnail outfit before firing up the machine algorithms. The technology takes into account more than 50 elements of each user’s personal attributes—everything from general fashion preferences to specific body shape—to make personalized recommendations, according to Fast Company.
Update 08/18/2018: Threads took in an additional $20 million from a funding round led by C Ventures to hire more stylists, and also more engineers to optimize their technology and expand their business. This brings the company’s total funding to $36.3 million.
Stitch Fix is kind of a cross between food delivery services like Blue Apron and Coco Chanel (whoever that is). The San Francisco startup has collected about $46.75 million in investments, with the last round, a $30 million Series C, coming in June 2014. The basic business model is subscriber based, where customers fill out profiles and pay a small fee to have a human stylist pick out five pieces of clothing that are delivered right to the customer’s door. You only pay for what you keep, and shipping is free. The company relies heavily on algorithms for all facets of its operation. It has even let machines design clothes with some HI input; all three shirts sold out. You can geek out on a tour of how Stitch Fix uses algorithms for AI fashion here.
Update 6/23/2018 – They since went on to have an IPO which we talked about in an article titled “Is Stitch Fix an Artificial Intelligence IPO?“
Developing Snapchat-like filters for the fashionista, ModiFace is a Toronto-based company backed by $4.25 million in equity funding. Its facial modeling technology—also powered by Nvidia chips—helps customers save time exploring makeup and hair products through augmented reality-based web and mobile apps. Its platform is reportedly used by more than 150 sites, including names like L’Oreal and CoverGirl. You can download the ModiFace Mirror app or go to the Sephora Virtual Artist tool to play with eyeshadow and lipstick online to give you an idea of what you might look like as a drag queen.
AI fashion offers yet another glimpse into the future of how artificial intelligence will affect industries once thought largely immune to large-scale technological changes. It’s particularly interesting because it represents a cross-section of e-commerce, design and related fields like augmented reality. Now back to watching Zoolander.