ReadingStar 2.0

ReadingStar 2.0 is an offline AI-powered educational app using music to support the speech development of autistic children, developed as part of the UCL Industry Exchange Network programme alongside Doruk Ersoy, Sinan Sensurucu, and Marko Stefanov. It was created in collaboration with Intel, IBM, the National Autistic Society, and MotionInput Games.

My contribution to the project was designing and implementing the user interface and user interactions, establishing frontend-backend communication, and developing the storage architecture for generated songs and settings.

I created the visual theme and the overall design language of the app, hoping to make something consistent, accessible, and pleasing. I designed the user flow and the structure and operation of each page, the most complex of which being the Play Song page: an intricate JavaScript mosaic of multiple simultaneous timers and dynamically updating HTML elements producing a synchronised karaoke-style experience. I implemented RESTful endpoints using FastAPI to enable robust frontend-backend communication, and designed the data models and storage system for managing instrumentals, lyric sets, and user configurations.

The project website gives a more detailed look at the UI design and the implementation of ReadingStar 2.0.

I do not usually endorse the use of AI to create music, however, this university project was developed solely for educational, non-commercial purposes, designed as a fun tool for children who struggle with their speech (and our generation is not going to win any awards).