Development of a Generative AI-Supported Reflection System for Sustaining Learners’ Motivation

Abstract

In the era of rapid reskilling, adult self-directed learners often struggle with volitional persistence in isolated digital environments. This research addresses these challenges through the design of an AI-driven "Learning Reflection System" grounded in Instructional Design (ID) principles. Developed using Gemini Flash-2.5 and Streamlit, the system integrates Goal-Setting and Reflection Chatbots supported by iterative templates. An expert review validated the system’s theoretical integrity and educational utility, confirming high alignment with the ARCS-V model for domainindependent guidance. This study establishes a foundation for integrating Generative AI with ID theories, offering a scalable approach to personalized coaching that facilitates persistence.