T-Boba

Happy
Subtle
Complicated
Dumb dumb

T-Boba

Welcome to our special T-Mobile branch: T-Boba! Here, customers interact with our sample mobile devices to enjoy the awesome T-Mobile experience! As they play with our devices, their user inputs are captured and analyzed which provides insight into what device we can recommend, what data plan they'll likely choose and much more!

As an example of what kinds of functionalities can be implemented, we developed a simple web application which uses the device's front-facing camera to analyze our customer's facial expressions. Based on different facial expressions, we provide the sales representatives with information to help them improve the customers' experience or recommend discount options to the customers.

Further Improvements

We can integrate various other user input data:

Apps they interacted with and dwell time on each app
Motion sensor data to see if the users turn the device around to check out the back look.
Front-facing camera to detect the facial recognition: expression, age, gender, etc.

We can try to answer these questions:

Are the customers happy with this device?
Are they likely to purchase this device? Or maybe this model but with other specs.
Are they more likely to use data or phone or voice chat? And thus what data plan best suits their need?
What apps and app layouts interest the customers more.

The work flow of a completed version would be: a machine learning (with models such as deep neural networks) service deployed on Azure will be continuously running to receive queries with facial and other aforementioned features detected and recognized using Azure Face API called by a mobile or web app. Upon receiving the results from the learning model, the app will react accordingly to lead the customers to the optimal next-step. Until the reaction, the app is only run in the background and thus not visible to the customers.

Actual personal information only exists in the intermediate level. Therefore, there shall be no privacy concern.