A.R.O. Reveals Saga, An “Ambient Companion” That Watches What You Do To Make Personal Recommendations

We’ve just been given a first look at Saga, a new mobile companion emerging from Seattle startup A.R.O. You can think of Saga as Siri’s little sister, perhaps. Instead of asking it questions or giving the app simple tasks (what’s the weather, add meeting calendar, e.g.), Saga is there, quietly tracking your behavior, your location and learning about your preferences, in order to make smarter recommendations about what you should do next. It’s the next evolution of those “ambient location” apps which were all the rage at this year’s SXSW, perhaps.

A.R.O is run by CEO Andy Hickl, formerly of Swingly, the NLP-based answer engine launched back in 2010. Saga is not the company’s only product – A.R.O. currently offers Bubbleator, too, which is a live wallpaper for Android. You may also remember hearing of A.R.O. back in 2010, when it first launched a suite of “semantic” core apps for Android phones backed by Microsoft co-founder Paul Allen. Those Android apps have since become part of Xiant Software, a Paul Allen company and a brand that Chris Purcell runs. They do email and productivity apps for the Kindle fire right now, with more to come. Saga is A.R.O.’s new direction, run by an all-new team.

Saga: The App For Your Past, Present & Future

Saga is a somewhat complex product to describe. At its core, it’s aiming to deliver value in return for the abandonment of our privacy through location-sharing. That’s something which the new-ish “ambient location” apps have been struggling to do. For Highlight, Sonar, Banjo, INTRO, and the like, location-sharing has served one purpose: networking, both business-related and/or personal. That’s not enough of a return on the investment (namely, battery drains), for some uses, however.

If anything, Saga is more of a competitor to Foursquare, the local discovery app that recently shifted focus away from the check-in to focus instead on its recommendations engine. Foursquare was able to do this because it had finally amassed a large enough place graph tied to our social connections and history of our visits to make sense of that data on a personalized level. Saga will attempt to do much of the same, but in a different way.

There are three components to Saga’s app: location-tracking, recommendations, and gamification. It’s hard to not see the Foursquare comparison, given those elements. However, unlike Foursquare, which still relies on the manual check-in to record your location, Saga will automatically locate you – and you don’t have to open the app for this to happen. But if you do open the app while at a venue, Saga will tell you things like how long you’ve been there, how many times you’ve visited and it will even guess at what you’re doing, which you can correct when wrong. And it will make recommendations as to what to do next. (You’re at Home Depot? Maybe you should grab a bite at that Indian place up the road, after all, it’s lunchtime, you’re hungry and you love Indian food.)

With all the data it collects like this, Saga can later provide you with a history of your activity for the day, week, month, etc. It can show you where you’ve been, how far you traveled, your top places and more. Basically, all that “quantified self” data that some people get a real kick out of.

Recommendations Are Key

But Saga’s bigger draw will be it’s ability to get to know you and make recommendations. At least, that’s the claim. And these won’t necessarily be places to eat, drink or shop, as in Foursquare, but other things, too. For example, Saga could learn your commute and warn you about traffic. It could remind you that it’s been three weeks since you took your dog to the dog park. It could recommend a whole day’s activity based on a theme (foodie, health nut, etc.). It could tie your bookmarked venues to those of your friends and recommend you get together soon and where.

Not all these features are live, however – they’re Saga’s potential. Of course, how well it works will be based on a number of factors, including the accuracy of its recommendation algorithms, its location accuracy, the overall users experience, and more. But the location accuracy seems promising, because it’s not based on GPS only. Instead, Saga uses a combination of factors (past history, similar users, accelerometer reads, set preferences, routines, etc.) to determine where you are. (The platform appears to use Alohar Mobile, more on that here, but the technology was built in-house, we’re told.)

The app is scheduled to launch in late June or early July on iOS, with Android to follow. More details are here.