How AI Is Set to Make Online Shopping Even More Tempting
WHY YOU SHOULD CARE
Because Diane Keng is fixing your browsing experience with artificial intelligence — then conquering the rest of the world.
By Molly Fosco and Sean Culligan
A few weeks ago, you found the perfect Father’s Day gift — a T-shirt of your dad’s favorite band. You bought it online and had it shipped to his house. This week, the same online store keeps suggesting Phish T-shirts, size large, instead of the women’s athletic tops you’re actually looking for. The algorithm can’t distinguish between shopping for your dad or yourself.
Not if Diane Keng has her way.
Keng, 27, founded San Francisco-based Breinify in 2016 to bring the power of artificial intelligence to online retail. Breinify uses temporal data — data that varies over time like weather, traffic and user behavior, typically used by transportation systems such as air traffic control — and ties it together from multiple sites to help brands better understand and target their customers. “We were purely surprised this wasn’t already translated into online retail,” says Keng. “At the end of the day, how you behave last week might not be the same as this week.”
But retail is just the beginning for Breinify. Keng says joining temporal data with AI can predict outcomes in ways that transform everything from pharmaceuticals to fintech to agriculture — likely within the next decade.
Traditionally, customer analytics have been based on past behavior rather than predictive …
It’s no surprise Keng wound up leading a tech company. After moving to the U.S. from Taiwan at age 7, she grew up in Cupertino, also home to the Apple campus in the heart of Silicon Valley. Her dad worked in venture capital and her mom left a career in finance to be a homemaker but was constantly encouraging Keng and her older brother. “My brother and I grew up understanding that there are a lot of new things in the world and you can either be making them or consuming them,” she says. Keng chose to make them.
Breinify is Keng’s fourth business. She’s been an entrepreneur since high school, including a business she launched with her older brother called MyWEBoo, a platform for managing social accounts that earned a spread in the Wall Street Journal. The siblings sold the company (she wouldn’t reveal the amount) and then it was off to Santa Clara University for Keng — a 15-minute drive from where she grew up — to get her engineering degree. “I had a lot of ideas I really wanted to build and by learning how to code in my first year I could build things without being dependent on other people,” Keng says.
After college, Keng did stints as a software engineer at Apple and Symantec but always intended to lead her own business again. At a conference in San Diego, she and her future co-founder, Philipp Meisen, hit it off after she spotted him at the hotel coffee shop taking up all of the available outlets. The two quickly found that their research intertwined incredibly well, they had the same intense work ethic and a passion for making a difference. “I realized Diane always has this visionary view on things,” Meisen says. “She sees beyond just a simple solution.”
Four years later, Keng, Meisen and their team of eight have built an incredibly powerful tool. Traditionally, customer analytics have been based on past behavior rather than predicting the future. But Breinify uses AI to make inferences from the raw web activity it receives. For example, if you use the Wag app and go to PetSmart once a month, it can infer you have a dog. It also understands that you’re interested in dog toys on Friday and you restock the dog food on Tuesday, and it will feed you information timed accordingly.
Breinify works with major brands like Hershey, BevMo and Hidden Valley Ranch and gets access to the data that customers have already opted into by visiting these company websites. The engine makes predictions about customer behavior that these companies could never do on their own without an extensive in-house data science team.
Keng emphasizes that Breinify never shares the data they access and they don’t sell data for profit. But privacy concerns are inevitable. “You can’t make predictions about people without knowing about people,” says Jacob Furst, a professor at DePaul University and an expert in AI, security and privacy issues. The higher your moral ground on privacy, the less useful your predictions will be.
Keng is quick to point to the sunny side of big data. In the financial sector, for example, Breinify’s AI could be used for fraud detection because the algorithm can spot suspicious activity. In farming, Breinify’s ability to incorporate weather data could save crop yields and improve operations. “The key takeaway is that many industries and pain points can benefit from a lightweight, easily scalable AI that can work toward custom goals,” Keng says.
As a young female founder with four companies under her belt, Keng also works to help other women advance in the tech industry and entrepreneurship. “I do fundraising walk-throughs for founders and a lot of it is about shifting the mindset around how women think,” she says. When it comes to founders and CEOs, “the biggest differentiator between women and the typical White man is that women are not able to talk bullshit as well — they second-guess themselves and put themselves down.” Keng encourages female founders to practice speaking in front of a mirror confidently, something that’s helped her in the past.
Outside of work, you can probably find Keng at one of San Francisco’s many museums or outside on her balcony tending to her recently purchased tomato plant. “It’s massive!” she gushes.
As is the power of her innovation.
Correction: This story has been updated to reflect that Breinify launched in 2016 (not 2015) and to clarify that Keng was talking about the differences between male and female founders and CEOs, rather than men and women generally.
OZY’s 5 Questions With Diane Keng
- What’s the last book you read? Purple Cow by Seth Godin. It’s about how to spread the message you’re focused on. I work with a lot of C-level execs and to do that you have to think like an old man.
- What do you worry about? With all the buzzwords around AI, I worry about the ethics and morality of what’s happening. How AI is used relies on who’s creating it.
- What’s the one thing you can’t live without? Probably my dog, Thor.
- Who’s your hero? My mom, and my dog’s namesake, the superhero Thor. My mom sacrificed her career to stay at home for me and my brother. She wasn’t an overbearing mom and she always made us think about our options. She did a good job balancing between the emotional and logical.
- What’s one item on your bucket list? I would like to do a trip from Somebody Feed Phil, where he [Philip Rosenthal] goes all throughout Southeast Asia and tries food from everywhere. It looks delicious and he video chats his parents at the end, which is hilarious.