AI を活用した事業企画プロセスの効率化に向けた仕様駆動開発型プロトタイピング

抄録

Generative AI has accelerated early-stage prototyping in service and software development. However, when AI is used without explicit specifications, generated artifacts may appear functional while deviating from user needs and design intentions. This paper proposes an AI-catalyzed HCD–SDD integration workflow that transforms qualitative user intentions obtained through Human-Centered Design into structured specifications for Spec-Driven Development. The workflow consists of four phases: intention extraction, intention structuring, specification conversion, and iterative generation, evaluation, and improvement. An exploratory case study was conducted in a service design and development project for a professional assessment task in a financial context. The results suggest that the proposed workflow can connect HCD-derived insights to requirements, specifications, and implemented prototype functions. The iterative improvement loop also showed a reduction in estimated improvement cycle time and an increase in System Usability Scale scores. These findings indicate the potential of the workflow to support prototyping while maintaining consistency with user intentions. The study is exploratory and based on a single case.