Full TitleAccelerat.ai: Multi-lingual Conversational AI with Modern Contact Centers
Conversational Al solutions have the inherent ability to offer an increased level of accessibility by being available through text or speech-based interfaces. This caters to the needs of different target groups, such as the blind and visually impaired, that benefit from speech interfaces, or as the deaf and hard of hearing, that benefit from text interfaces. Other groups, such as persons with some cognitive disabilities, will benefit from the possibility to use both speech and text interfaces simultaneously. Nevertheless, conversational Al solutions raise potential barriers for users with disabilities, some of which will impact even the user experience for everyone, regardless of their abilities.
For instance, by delivering messages in sequence, they raise issues about the message content length and pacing for users with different types of needs. Another example is how the communication style affects users. Conversational agents exhibiting empathy can improve human mood and therefore the customer experience. This introduces challenges and responsibilities to ensure that the conversational solution does not negatively impact users emotionally or psychologically.
Thus, this project encompasses the development of an automated accessibility checking solution for conversational Al systems. In order to ensure such development, the present activity will include major research tasks such as (i) requirements specification, (i) design and prototyping, (ill) user evaluation, (iv) pilot implementation.