Practising public speaking plays a major role in language training. Besides the interactional dimension, it requires the promotion of a set of verbal and paraverbal skills. Within the domain of paraverbal skills, the prosodic dimension is particularly important. Prosody traits, scientifically studied by Phonetics and Phonology, as disciplines within the area of Linguistics, can include features related to speech melody, speech dynamics and voice quality (Cagliari, 1992). As far as speech melody is concerned, aspects such as intonation are relevant because of the semantic and pragmatic clues they convey. As for speech dynamics, pauses play a major role in determining the rhythm of speech. Finally, with regard to voice quality, issues such as volume and register stand out. In oral communication, orthoepic competence (Pérez, 2021), investigated by the same scientific branches of Linguistics due to its affiliation with prosodic competence (Souza & Pereira, 2019), is equally important as it contributes to the transmission of authenticity in language use. In higher education, in the context of specialised training for business communication professionals and in the training of interpreters (Penha, 2015), these skills play a central role. Accordingly, in this study, we intend to critically analyse the contribution of Artificial Intelligence tools specialised in self-diagnosing, diagnosing and monitoring oral discourse - Speeko and Orai (Halimah, Lustyantie & Ibrahim, 2018) - to the development of prosodic and orthoepic skills in higher education students that require a confident public speaking ability.

Keywords: Prosody, AI speech coach, public speaking skills, oral training, higher education.