<![CDATA[The advancement of artificial intelligence technology has accelerated the development of interactive and responsive voice assistant systems. One of the widely adopted technologies for speech recognition is the Whisper model, which is capable of automatically converting speech into text. This study aims to implement the Whisper model in an AI-based chatbot voice assistant system designed to support access to academic information through voice interaction. The developed system integrates Whisper as a speech-to-text module, an AI chatbot for processing user requests, and text-to-speech technology for generating voice responses. System evaluation was conducted using the BlackBox Testing method, covering user authentication, voice assistant access, voice recording, and audio processing to generate voice responses. In addition, the Whisper model was evaluated using ten Indonesian-language test sentences. The results showed that all main system functions operated according to the specified requirements. The Whisper evaluation also demonstrated that all test sentences were successfully transcribed in accordance with the given inputs under the conducted testing scenarios. The findings indicate that the integration of the Whisper model, AI chatbot, and text-to-speech technology is capable of producing a responsive voice assistant system that supports natural voice interaction. Therefore, the developed system has the potential to serve as an educational support tool that facilitates access to academic information through voice-based communication.Keywords: Voice Assistant, Whisper, speech recognition, chatbot, voice interaction, BlackBox Testing.]]>
Copyrights © 2026
