“Sort of” an Awesome Contexual Search

Posted by Jeff Lu on March 6, 2009

“I’m Jim Heising and I’m with Sortuv a company.”

A well rehearsed line from the co-founder of Sortuv which was mildly amusing but was probably a bigger hit with the (unfortunately under-represented) female crowd at our a weekly Seattle start-up happy hour (Hops and Chops) table last night.  Hops and Chops occurs every Thursday at Linda’s Tavern in Capitol Hill and it has become pretty well known around the community as a great place to meet people from the local start up community in a low-key setting.

Last Thursday evening was the first time I was able to escape my cubicle and head over to the weekly networking event.  My favorite part about my job is meeting new people and companies and being blown away by their new ideas.  I didn’t suspect it right away but I was about to be blown away by Jim and Sortuv.

(click to download) iPhone Screenshot: Reviews, rating and "Sortuv similar to" suggestion

Sortuv’s has a product called “near+now” out and it currently only runs on the iPhone and Facebook platform as an downloadable application.  Near+now is a location specific personalized search platform for local restaurants and events.  Jim gave me a quick demo at the bar and my first instinct was to think that near+now was another Yelp, UrbanSpoon, NearMe or CitySearch, but I quickly realized that Sortuv’s app was special.  Like all the other restaurant/event apps, it allows you to rate restaurants you’ve been to and share your experience as an editorial, however that’s where the commonality ends.

Near+now is really an intelligent search engine being directed at the restaurant/events arena.  The application recommends restaurants in your area that you might like based on your previous ratings of other restaurants, but they use a complex and amazingly effective algorithm to do it.  Typically, search algorithms recommends restaurants based on direct and simple relationships like the cuisine, price, neighborhood, etc.  The problem with that keyword search and recommendation process is that it has a difficult time with finding things that are subjective in nature.  An “authentic” Chinese restaurant to my Jewish friend Adam is much different than what I consider to be an “authentic” Chinese restaurant.

Sortuv: Changing the way we search by finding relationships instead of key words

Near+now recommends restaurants based on millions of relationship queries to determine discrete similarities to other restaurants we like.  So instead of searching for “authentic” Chinese restaurant, Adam and I can each find our own special degree of authenticity based on other Chinese restaurants that we’ve liked.  Another example that Jim used was, hypothetically, if I really liked a restaurant in Chicago that is owned by a specific chef and I was in Seattle for work, near+now might recommend a restaurant in Seattle that is also owned by that chef.

It was at this point in our conversation that things really got exciting.  Currently, Sortuv’s algorithm is designed for finding relationships with restaurants to give you the most relevant answers to your search.  Jim asserted that with just a few months, Sortuv’s search technology can be applied to other topical arenas as well, such as clothing, consumer products, music, movies etc.  Sortuv’s search adds context to your search while sidestepping the furtive “semantic search” holy grail.  The potential to for Sortuv’s search technology to scale across multiple topical platforms is an exciting and I’m eager to see what the team has to us next.

Sortuv ‘s team consists of 4 technologists, each boasts varying degrees of success with previous startups:  Ron Franczyk, Jim Heising, Andrew Newman and Kalid Azad.  The Company currently has no institutional funding and its current revenue model falls into 3 categories:

1) Licensing an API for 3rd party developers to utilize Soruv’s technology and data

2) Targeted and location specific local advertising

3) Affiliate revenue

Clearly, the goal is to build the brand and to hit a tipping point with users interacting with the company’s data and search technology.  The Company currently only has apps for Facebook and the iPhone but is exploring other platforms such as Blackberry Bold and Palm Pre (when it’s released).

I hope I’ve done Jim and his Company justice with my coverage.  I’ve probably botched it so check out an interview of Jim that I’ve found on YouTube: