Università accessibile

As a research fellow at the University of Trieste from October 2018 to October 2019, I was tasked with developing a prototype voice application for smart speakers.

My goal was to design and develop a voice application that could help students and visually- impaired employees access their enrollement data and book exams sessions in a more intuitive way.

• Research: I conducted thorough research on current voice technologies and their capabilities. I evaluated two alternatives: a native voice interface using smart speakers and a chatbot created with the Botkit framework. After careful consideration, I selected Dialogflow as the preferred voice technology due to its versatility, cross-channel integration, and ease of use,
• To ensure a natural and informative user experience, I analyzed a preexisting GUI interface and identified the key functionalities that students would need to access. I developed an endpoint that acted as a webhook on the Heroku platform, allowing each intent to convey a parameter representing the requested operation and consume the associated university endpoint,
• I then proceeded to design and develop the voice application using Dialogflow and its API v2. The application architecture consisted of a front-end of approximately twenty intents in Dialogflow, connected via fulfillment to a webhook that consumed the university’s web service exposed by the Esse3 APIs,

I conducted usability tests with 20 students and it turned out that the voice application successfully met the students’s approval. Though a smart-speaker, they felt they would speed up the process of booking an exam. In addition, it provided students with a convenient and efficient way to access information about their university career, such as exam schedules and grades.