Virtual Booth

In many English as a Second Language (ESL) classrooms, students face persistent challenges in developing effective speaking and interview communication skills due to low confidence, weak pronunciation, and delayed feedback mechanisms. Traditional instructional approaches often provide feedback only after performance, limiting opportunities for immediate correction and real-time improvement.

This study introduces VIVAAI-PRO, an enhanced framework developed from the original VIVAAI system, designed to provide real-time cognitive coaching during students’ speaking performance. Unlike its predecessor, which focuses on post-response feedback, VIVAAI-PRO integrates live voice input and artificial intelligence to guide learners through an iterative improvement process while they speak.

The framework was implemented among Semester 5 students from two different departments at Politeknik Seberang Perai: Jabatan Perdagangan (DPM5B) students enrolled in DUE50132 (Simulation Assessment), and Jabatan Kejuruteraan Mekanikal (DKM5A) students enrolled in DUE50032 (Mock Interview Assessment). The system utilises freely accessible tools, namely Google Docs Voice Typing for speech-to-text transcription and ChatGPT for AI-driven feedback and response enhancement.

A key innovation of VIVAAI-PRO lies in its ability to transform imperfect speech into meaningful learning. Even when pronunciation inaccuracies result in incorrect transcription, the AI interprets the intended meaning based on context, sentence structure, and common learner errors, and provides corrected and improved responses. This shifts the focus from accuracy dependency to meaning-based learning.

Findings from classroom implementation indicate significant improvement in students’ fluency, confidence, response structure, and reduction in hesitation after repeated practice cycles. Additionally, the framework reduces lecturers’ marking burden by enabling immediate, automated feedback.

VIVAAI-PRO demonstrates strong potential as a scalable, low-cost solution for enhancing communication skills in TVET and higher education contexts. By leveraging accessible digital tools, the framework promotes inclusive, real-time learning and has the potential to be extended to other domains such as presentation skills, reading, and general communication training.