Wednesday, September 24, 2014
Tilofy: Making Sense Of The World Around You Using Location Data
Story by Benjamin F. Kuo
What happens when you apply big data and computer science to digest all of that social data being fed into Twitter and other social platforms--and cross match that to location information? Los Angeles-based Tilofy (www.tilofy.com)--out of USC's Viterbi Startup Garage--is doing just that, helping to highlight what is going on around you, by absorbing all of that location data out there at any particular time and place. We spoke with Ali Khoshgozaran, co-founder and CEO of the company, to learn more about the startup.
What is Tilofy about?
Ali Khoshgozaran : Tilofy comes from Time, Location and Unify. Our mission is to organize all of the world's real time location information. My degree is from USC in big data. Prior to Tilofy, I had technical positions at Yahoo, Microsoft, and Samsung, ranging from development, to program management, to senior technology person on Samsung's smart TV platform. Over the last year or so we've created Watsup, the easiest way to figure out what is happening around you, at any time. Behind the easy to use interface is a powerful, real time search and discovery engine, which creates a news feed of all of the intersting things happening in your surroundings. If you're familiar with Shszam, the mobile app which scans and tells you who an artist is, we've created something like Shazam for the real world. You can press a button, and it scans the social media chatter and organizes highlights aroundyou, of relevant conversations around a particular location and particular time. So, the highlight could be a celebrity sighting, an accident, a New York fashion show, and so on and so forth. It's a very interesting way of bringing the world around you to life, in a very easy to use interface on your phone.
How did Tilofy come about?
Ali Khoshgozaran: Before we started the company, me and my partner had participated in the USC Viterbi startup competition. As you probably know, Viterbi is the engineering school at USC. About 100 teams participated in that competition, and 10 teams were selected to be part of USC's Viterbi Startup Garage program. That program is a partnership of the University, Kleiner Perkins Caufield and Byers, which is a top tier VC firm, and United Talent Agency, the premium talent agency. Those three organizations have very strong networks inventure capital, entertainment, and academia, and put together the incubator. From there, we raised $600,000 plus to date, and recently revealed our app on the app store
What drove your interest in this area?
Ali Khoshgozaran: During my five years at USC, my research was in the area of spatial and temporal information management. Basically, that is figuring out how to deal with massive amounts of data which is time and location sensitive. I spent five years researching interesting ways to organize data on time and location. At Microsoft, I was part of Bing Maps, which is Microsoft's response to Google Maps. I was lucky and honored to see two sides of the problem, both from the pure academic perspective at USC, as well as from the Bing Maps team. That was a completely different view of the same problem. I was able to put those together, taking my expertise in this area in both academia and industry, to build Watsup as it is today. We now have a team of nine engineers and scientists, with experience from companies like IBM, Demand Media, and Google, all part of our core engineering team.
Where are you getting all of the location information, and is there enough location data out there?
Ali Khoshgozaran: That's a great question. The app sources information from two sources in particular, Tweets and Instagram. There's also some other data sources we're ingesting and sourcing to users. For Instagram, a significant portion of Instagram photos are now automatically geotagged. There's lots of location there for video and images. For tweets, percentage of geotagged information is lower, and it is a fraction of actual tweets, but we have developed intelligence ways of inferring location based on the media conversation. We're also using web articles and other data sources, and cross referencing those with location information. You can download the app and see the depth of information now available. We're in four major U.S. Markets now, including Los Angeles, San Francisco, New York, and Las Vegas.
What would be most interesting to get a user to use this?
Ali Khoshgozaran: It's funny, as people have the idea that Twitter and Instagram are the fastest ways of getting good real time information. However, although it's arguably the fastest way to generate information, it's not necessarily the fastest or easiest to consume information. What we have done for the user, is used a patent pending algorithm to sift through north of a million data points a day, and thousands of conversations per second, to help organize and cluster highlights. We're effectively taken away what is the cognitive overload which happens when you try to go through hundreds of tweets, Instagram updates, and Facebook updates, and the rest, and try to get the context of what is happening. Watsup makes it extremely simple for you to get an idea of what is happening around you. An example of that is the UCLA flood. When that happened, one of our investors said it was very interesting, because he was sitting stuck in traffic on Sunset Boulevard and had no idea of what was happening, where to go, or what to search for. He just new something was going on. So, he launched Watsup, and the first thing it showed was a highlight of the UCLA flooding. With a single swipe, he was able to see hundreds of videos, photos, and tweets from people on the ground. Give the powerful way of organizing data, you can get a birds eye view of everything happening in a city, from arts and culture to sports, to things like where the best places are to take a picture of the sunset, where celebrity sightings are, where there are restaurant openings, and even things like accidents and fires. It's all there in one place, in one, easy, simple to use interface. Our users want to be in the know, and have a fear of missing out, and want to be the cool people who know what's going on in the city, better than everyone else. They've never had this amazing source for that kind of information.
Finally, what's next for your company?
Ali Khoshgozaran: This is just the beginning. We've got a grand vision of organizing information by time and location, and this is just the baby step. We've spent seven months developing the algorithms to automatically filter through the data and create those highlights, and now that we have our application running on the iPhone, we are shutting up and listening to our users, to get a deeper understanding of how they are responding to the app. This is just the beginning. Obviously, we have more feature updates for iOS, and are going to Android, and preparing a desktop version. However, obviously, the deeper value of this is not in a cute little iOS app, it's all the information behind it, and figuring out in-depth, granular, and actional information and insights from that data.