Fashion forward – Isthmus

Joy Tang spent seven years working in the fast-paced field of high-frequency stock trading after graduating from Massachusetts Institute of Technology. With a background in math, economics and information technology, she thought she would spend her career working on the technical, quantitative side of the finance industry.

While Tang was crunching numbers and analyzing complex algorithms, she found that social media became her “only tunnel to the rest of the world.” A lover of clothing and fashion, she shares outfit photos with her friends on Facebook, Twitter and Pinterest and follows bloggers for style inspiration.

“Deep down in my heart, I’m just like any other girl who likes to dress up and look pretty,” says Tang, 31, who grew up in China.

Her intersecting interests in fashion and technology hatched an “invention moment” for Tang. Scrolling through fashion photos is fun, but if you want to add a piece to your own wardrobe, it can be tricky to track down where to buy it.

“When I see pretty outfits from my friends, I really want to know where they bought it, but it’s really awkward to ask them — I don’t want to sound like a copycat,” Tang says. “I follow a lot of fashion bloggers and comment on their photos, but they have hundreds of comments, and oftentimes they never have time to respond.”

Her desire to be able to “shop any photo” led her to create Markable, which launches this week in the iTunes store. Users snap or download a photo, upload it to the app, and the app finds the exact item or some similar options, which can be sorted by price. Markable is connected to more than 800 brands and stores, and users can make direct, in-app purchases with credit card or PayPal.

“This app will really help busy people to buy what they want immediately,” Tang says.

The visual search technology is able to identify objects in photos using “deep learning,” a type of machine learning that employs algorithms and hierarchical models to process data. The same technology was used to create AlphaGo, a computer program developed by Google that in 2015 became the first computer to beat a human at Go, a board game more complex than chess.

“It’s pretty much in the field of [artificial intelligence],” Tang says. “Lots of people know about this kind of technology, but not many people apply it to fashion — probably because most [people in the field] are men.”

So when Markable brings up the top 50 suggested products based on a user’s photo, that list is compiled based on millions of calculations made on the back end, Tang says. The deep learning system is trained to instantly analyze color, shape, pattern and texture, but the actual grouping and presentation is more subjective. Markable spent nearly a year gathering feedback from users and product data from popular clothing brands.

Tang started the company in Chicago in 2014 and now works out of 100state in Madison with a staff of nine full-time and five part-time employees. Dicky Lou, a creative designer, made the move to Madison with Markable after meeting Tang at a tech event in Chicago and joining the team.

“Madison is a great city for startups,” says Lou, who also grew up in China. “There are talents and resources, and the people are very focused.”

Tang agrees that Madison is an up and coming city for young tech companies, noting that venture capital firms look favorably on Madison because of its reputation for yielding high-performance investments. Cost of living is low, the environment is less competitive, and the presence of UW-Madison and Epic Systems is beneficial.

“I think Madison is actually a lucky city for startups,” Tang says. 


Write a Reply or Comment:

Your email address will not be published.*