Co-founder and CEO of Visenze
YEARS IN OPERATION: 5
THE TECH: The development and application of artificial intelligence and image recognition technology to power visual search and product recommendations.
[dropcap size=small]I[/dropcap]n a scene from Minority Report, AI-powered billboards greet Tom Cruise’s character and pitch products to him as he walks past them. As he enters a clothing store, a bot recalls his purchase history and recommends similar pieces. While talking posters are not yet a reality, some semblance of such intuitive targeted advertising is already in motion. But few are aware that a Singapore start-up is behind this cutting-edge technology.
Founded five years ago, Visenze utilises deep machine learning based on image recognition algorithms to drive picture-based search and product recommendations on e-commerce sites such as Uniqlo, UK fashion site Asos, and Flipkart, India’s largest online retailer. This, co-founder and chief executive Oliver Tan says, is just a glimpse of retail’s smarter future. After all, the first forms of communication were cave drawings and hieroglyphs, so Visenze is bringing consumers back to their native instincts, he adds.
“Every retail marketer’s dream is to predict consumer intent. But they don’t know how to personalise, recommend and create search processes that link back to the searcher,” the 49-year-old says, as he demonstrates on his smartphone how Visenze processes an image of and accurately searches for a pair of ladies’ heels, before recommending similar shoes – all in under five seconds. “We can aggregate consumer data to understand products people are currently searching, and then influence how they shop and lead them towards action, which is what retailers want.”
Even as Visenze promises to lift retail out of the doldrums, Tan recalls the resistance he and his three co-founders faced when they started out in 2012. Few people in Singapore understood AI and the market rejected their idea for fear of upsetting the status quo. “People here kept asking us how we were going to do it, but the big question should be why. That made me determined to go to the US, the UK and Taiwan where there’s critical mass. Once you start talking with like-minded people, you discover how you can do things better,” says Tan, who sits on the invitation-only Forbes Technology Council.
A year after being in stealth mode, Visenze began working with Rakuten Taiwan in 2013 and word about the start- up’s capability quickly spread. Besides e-commerce, Visenze, which raised US$14 million (S$19 million) across two rounds of funding, is also utilised by clients such as wallpaper company Goodrich Global and intellectual property start-up Patsnap to solve search problems regarding inventory. Last year, Visenze processed over 350 million images. Retail clients saw search-to-purchase conversion rates improve by up to 80 per cent, Tan says.
In a way, Visenze’s revolutionary streak mirrors Tan’s rebellion on his career path. The corporate high-flyer ditched a lucrative position at a Nasdaq-listed multimedia company in pursuit of self-fulfilment, something he’d had a taste of when he joined home-grown cyber security provider E-Cop (now Quann) in the early noughties. “I was becoming too specialised in what I was doing. I wanted to discover myself more,” says Tan, whose three children then were under 12.
“I didn’t want to retire as a suit, and I hadn’t made a big difference.”
“People are like teabags. You never know how good you are until you are in hot water. Those five years at E-Cop were my best. I didn’t know how adaptable and resourceful I was, until I was thrown in the deep end. Then I returned to the corporate world because the pay was too good to reject, but I found myself slipping back to old ways. I didn’t want to retire as a suit, and I hadn’t made a big difference.”
Though the road towards a Minority Report world of instinctive advertising may be wrought with speed bumps, Tan firmly believes in AI’s transformative potential in everyday lives. Next year, Visenze is planning to roll out a video commerce experience that uses new AI visual recognition technology. It also intends to power chatbots with advanced image recognition capabilities to act as digital shopping concierges.
“I’m a long-distance runner. I have a lot of stamina. I believe there’s a lot that we have not accomplished. What drives me on is the unfulfilled task ahead.”
IN 60 SECONDS
I’ve learnt that it’s not always about having the most original idea, but being better than the original concept.
Disruption to me is… I prefer to use “transform”. When you want to disrupt, some people have a negative reaction. Transforming is a better conversation to have.
My modus operandi is fail fast, fail small. Don’t make big expensive mistakes that can kill you and make you risk-averse. It’s the iterative process that takes us forward.
Browse previous Power Lists:
CLOTHES WOOL AND VISCOSE BLAZER JACKET, WOOL SWEATER, & WOOL AND VISCOSE PANTS, FROM HUGO BOSS GROOMING CHRISTINE LIM, USING SHU UEMURA