Saga: A personal assistant who strives to be a prophet

In recent days, according to media reports, a personal assistant named Saga to the Apple App Store ready to grab a job with Siri. The trick of this app is to tell users what they need to know when they need to understand. In a large number of data and algorithms to support the application can be in the user out of the time to know their intentions, to achieve the foresight, to provide users with surprise services.

Saga is developed by A.R.O. Andy Hickl, its CEO, does not think saga is equivalent to Siri because Siri will only answer questions that the user offers. With the user agreeing to keep track of location information, Saga can learn to know where and how long the user will be, and how long it will take to get home now. Near lunch, it will recommend a good place to eat according to where the user often goes. It also tells the user how far the trip has gone and classifies the places he has visited according to their length of stay.

In the words of Hickl, Saga is equivalent to what the user often does, a quantized self acting (that is, mastering its behavior).

Pattern training is a must.

Of course, the accuracy of this prediction and the accumulation of data is closely related, so saga also need to go through a period of time to learn to know yourself.

Especially in densely populated urban areas, it's hard to find out if you're in a café or a restaurant across the street or your own apartment. Saga's approach is to set the initial setting according to the popularity of the site. But after 48-72 hours of feedback (telling the application where you are), the application can master your line and make more accurate predictions.

Another need to spend time learning is the user's intention, the current situation is not so advanced means can directly know what users want to do.

For example, when someone enters the Walgreen at 11 o'clock in the evening (Fortune 500, the food and medicine chain), he may buy a beer for the party and possibly buy some medicine for his sick wife. In this case saga will be able to mark both possibilities, but if you know that this person usually rarely goes out after 9 o'clock in the evening, Saga will directly choose the second possibility.

The more data, the more useful

Saga future plans to become a platform for other applications so that saga can master more user data. Even A.R.O. It is also envisaged to incorporate data from user friends to provide more precise advice. The more data is associated, the greater the application will be.

In addition to mobile phones, Saga will likely be deployed to new devices such as wearable display devices.

Of course, Saga will also face competing applications like Google Now. The latter has a natural home advantage in the Android Market, but Saga also has two points to make, one that protects privacy from design and that does not sell user data, which can also help saga gain moral advantage.