Palo Alto-based Ness Computing launched its dining recommendations app only about three months ago and already it’s being wooed by the likes of Goldman Sachs, was just listed in the Staff Favorites section of Apple’s iTunes App Store, and closed a Series A funding round with some big time investors.
Did I mention it’s a dining recommendations app? I was skeptical at first, too.
Sure, you can get a good read on which restaurant has the best Pad Thai using Yelp or Google Places, but Ness is different because it’s smarter. And it should be—the iOS app was created by a team of engineers, programmers, and machine learning Ph.D.s.
It doesn’t just use your location to suggest nearby restaurants. It wants to get to know you first.
When you launch Ness the first time it asks you to rate 10 restaurants in your area. I gave McDonalds one star, Subway two stars, the Mexican joint my husband and I like four stars, and so on.
It also has you sync your Facebook and Foursquare accounts to the app, although that’s optional. Once you do that, it can see which restaurants your friends recommend.
Although only four of my friends are active enough on social networks to count in Ness’s algorithms—meaning they’re using Facebook or Foursquare check-ins—I can now see that Jason F. likes a place in San Francisco called A 16.
But here’s where it gets interesting. Ness says I’ll dislike that particular place.
Just how does Ness know I won’t dig a pricey Italian place in SF? It has to do with how I’ve rated other restaurants and so far it only knows I frequent cheap eateries.
As I give it more information, however, it will not only use my past recommendations to give me better ones but it will also use natural language processing and social data mining to pull my friends’ data into the equation. The result is an app that learns to provide me with personalized search results.
Right now it’s only for restaurants, but Ness cofounder and CEO Corey Reese says he plans to expand the service to other verticals such as concerts and venues, music, and shopping.
His goal is lofty. Just as Google has indexed all the objective information in the world and put it at people’s fingertips, Reese plans to give people answers to all of their subjective questions: What should I watch on TV tonight? What Tom Cruise movie should I see this weekend? What restaurant should I eat at for lunch today with the guys at work? What shoes should I buy here at Nordstrom? Where should I travel?
To do that, he wants to build what he calls a “listening tower” that taps into all of the different places in which people store information about how they spend their time and money. While today Ness relies on user input and social media behavior to provide users with restaurant suggestions, someday it could use credit card transactions as well.
In essence he wants to see Ness become like Google, but more personal; like Siri, only smarter.
The brains behind Ness’ augmented intelligence include Nikhil Raghavan, formerly with Yahoo’s structured Web search team; Paul Twohey, a former geospatial search engineer at Palantir; and Steven Schlansker, a programmer from UC Berkeley. Reese has also gone through pains to hire several Ph.D.s in machine learning including Jeremy Schiff, who was previously co-founder of Fotoflexer and Scott Goodson, a former senior engineer on Apple’s iOS team, reports Forbes.
Ness has raised $5 million in Series A funding from Alsop Louie Partners, Khosla Ventures, Bullpen Capital, Tomorrow Ventures and angel investors such as Palantir Technologies co-founder Joe Lonsdale. (Palantir’s computer scientists create highly sophisticated data-sifting and visualization tools that help three-letter government agencies, banks and other organizations conduct cybersecurity, counterterrorism, and fraud detection.)
For now, you can only use Ness on Apple devices.