Teacher as Toolmaker: Developing AI-Assisted Oral Assessment in Mathematics

Abstract

 This study examines the phenomenon of teacher as toolmaker through the development of an AI-assisted oral assessment platform in mathematics education within a post-pandemic context. Addressing the misalignment between educational technologies and classroom realities, the study adopts a qualitative instrumental case study approach using development artefacts, AI interaction logs, and classroom implementation notes. The findings identify three key dimensions. First, pedagogical primacy ensures that instructional goals drive technological design. Second, a symbiotic human–AI collaboration emerges, where AI supports technical development while the teacher retains pedagogical authority. Third, the platform enables ecological integration by combining task delivery, response capture, recording, and analysis into a unified system, enhancing both efficiency and insight into students’ reasoning. The study demonstrates that teacher-led innovation can produce contextually responsive digital tools. It concludes that AI should function as a pedagogically grounded support system that extends, rather than replaces, teacher capacity, with implications for teacher development and educational technology design.