hackabletype

Nate Waddington's Portfolio

Triple-Checked

In an era of big data, there’s no hiding
for the small print

“Watson’s visual recognition, natural language processing and machine learning combine for a personalised, interactive AI introduction. Watson reads thousands of pages of T&Cs from the most downloaded apps, extracting the relevant privacy data, pinpointing the risks, before suggesting ways each guest can protect their invaluable information.”

– AKQA Triple-Checked


Triple-Checked is a part of the Watson Activations project. Watson Activations is a Kiosk system built for IBM that allows for multiple and varied experiences to be installed and shown around the world at various conferences and trade shows. Over the course of two years, several activations have been built for the Kiosk system, and it’s design ensured that these activations, once built, can be run without a specific need for AKQA staff to be involved (or even to know about the event). The Triple-Checked activation, as one of the launch activations, is a little different in that it has a camera option, though any subsequent activation can use it, it is thus far the only one that does so.

Roles and Objectives

  • Setup the IBM Watson machine learning system to recognize 20 top used app icons
  • Helped write the software to interface with the camera
  • Worked on the middleware to query Watson with images of phones
  • Worked with vendors to design physical enclosure of the camera in the Kiosk

March 2018

Filed under AKQA, programming, prototyping, production, installation, artificial intelligence, machine learning, Watson