Singapore’s ViSenze raises $10.5M to bring the benefits of AI to e-commerce

Next Story

Official CPSC recall goes out for Galaxy Note 7 — replacements will be available September 21

There’s another round of funding for an artificial intelligence startup in Asia that’s focused on e-commerce. Fresh from a Sequoia-led investment in Mad Street Den last month, now Singapore-based ViSenze has announced a $10.5 million Series B raise.

Existing investor Rakuten Ventures led the round alongside WI Harper Group and Enspire Capital. A number of other investors also participated, including SPH Media Fund, FengHe Fund Management, Raffles Venture Partners, Phillip Private Equity, and UOB Venture Management. The startup was spun out of the National University of Singapore more than three years ago and it raised a $3.5 million Series A in February 2014.

ViSenze develops artificial intelligence for use in e-commerce, working with companies like Myntra in India and Zalora in Southeast Asia. Its first product was a visual search feature, and there are two new releases coming soon that will expand its focus to video and help retailers with off-site discovery, ViSenze CEO and co-founder Oliver Tan told TechCrunch.

“We believe shopping behavior is very inspired by enriched content in the social web and media space,” Tan said in an interview. “If we can find interesting connections and help leverage content that inspires shopping behavior, then that can drive more engagement and conversions.”

Currently being tested in closed pilots, one of the new products will enable contextual advertising inside and around video content. That’s to say that content owners will be able to insert relevant advertising into their videos, while advertisers will be able to contextualize their web advertising by making it relevant to the videos that they appear against or near. Global video site Viki, which is owned by Rakuten, is one partner — you can imagine that fashion-based advertising would resonate well against Korean dramas, one segment of programming that Viki is known for. So, for example, a Korean drama could include in-video ads for outfits worn by the cast, with links out to e-commerce stores that sell them.

The other new feature will enable image to text. That will allow for more customizable searches from image uploads, for one thing. Currently, shoppers can upload a screenshot to find related products in an e-commerce site, but the image to text feature would enable them to customize their search for easily. That could mean selecting a different color, shorter/longer sleeves, etc.

More wider, Tan said that the funding will be used for hiring — he expects to “easily” double the current headcount of 40 before this year is out — and to “double down on R&D and innovate further.”

“The funding will accelerate our go-to-market strategy, and a large part will be re-invested in our tech,” he added.

In addition, ViSenze intends to grow its offices in New Delhi, India, and San Francisco, which are primarily for business development, and open similar outputs in the UK and China. Tan said he is spending increased time in the U.S., where he believes there’s a heightened level of understanding around the benefits of AI and the importance of visual search for commerce.

“It’s a very different market from Asia with different levels of sophistication. Many people [in the U.S. retail space] just get us right away,” he explained.