Speakit

A screenshot from the Speakit app, showing the main camera UI with three objects highlighted and labelled in Danish: Ur, Mikroovn and Køleskab. Underneath is the text "Tap on an object. Get close to the object. Make sure it is well lit.Hold the camera still until you see the yellow frame." Below are two buttons, from left to right: a button with a left-facing arrow and the label "Back", an outlined button with an armchair silhouette and the label "Home".

About

An AR mobile app that uses object recognition AI to allow the user to learn Danish vocabulary in everyday contexts, by scanning their environment. It was developed at TimestoryVR for the Danish Red Cross (Røde Kors).

Technologies:

  • Unity
  • Object recognition
  • AR
  • Machine translation

My role

My role was as a generalist - to help the team integrate new features and client changes, into the existing codebase.

Applying an MVC architectural pattern, alongside a message broker implementation, I decoupled components and API endpoints from the existing UI to enable those resources to be used in the new features, without adding duplicate effort and extending development time. This helped us to address more bugs than expected in the time we had.