From Education to Evidence: A Collaborative Practice Research Platform for AI-Integrated Agile Development

Abstract Agile software development evolves so rapidly that research struggles to remain timely and transferable—an issue heightened by the swift adoption of generative AI and agentic tools. Earlier work has highlighted a theory gap, where findings lack a clear conceptual basis for reuse, a time gap, where results arrive too late for practical decisions, and a transfer gap, where insights are not packaged in forms practitioners can readily apply. This paper introduces a project-based, AI-integrated agile education platform as a collaborative practice research environment positioned between controlled studies and real-world industry settings. The platform enables rapid inquiry through sprint rhythms, quality gates, and authentic stakeholder involvement. We present a framework that specifies iteration structures, recurring events, and quality gates for AI-assisted engineering artifacts. Early results from several semesters, including project pipeline development, cohort growth, and stakeholder participation, suggest that the platform can generate practice-relevant evidence efficiently while preserving the contextual conditions needed for reuse. Finally, we outline future steps to enhance governance and evidence capture.