The need for user-centered and agile solutions has driven the adoption of methodologies such as Design Thinking in knowledge engineering projects. While the Design Thinking emphasizes empathy, iteration, and innovation focused on user experience, the knowledge engineering aims to build intelligent systems based on formalized expert knowledge. Despite the conceptual alignment between these approaches, the integration of Generative Artificial into this context remains underexplored. This study proposes a model that incorporates generative AI tools into the design thinking process to accelerate and enhance the development of knowledge-based systems. Based on the Design Science Research (DSR) methodology, a narrative literature review and exploratory research were conducted to identify generative AI techniques applicable to each design thinking phase: Empathize, Define, Ideate, Prototype, and Test. A total of 17 generative AI approaches were mapped and organized into a model that supports small and agile teams in knowledge engineering projects. The model was instantiated through the development of GPT-based agents customized for specific tasks, such as persona generation, empathy mapping, requirements analysis, and prototype creation. These agents leverage prompts containing transcripts, observations, or interview data to generate detailed and realistic outputs, streamlining processes that are traditionally manual and time-consuming. One example presented is the Persona Generation Agent, which creates structured user profiles and illustrative images from simple textual input. This integration contributes both theoretically and practically by demonstrating how generative AI can be used to enhance user experience (UX) focused design processes in knowledge engineering. The proposed model promotes more efficient workflows, while keeping the user at the center of development. It also supports interdisciplinary collaboration, faster iteration cycles, and the creation of intelligent systems that are more aligned with user needs. Future work includes empirical validation of the model in diverse application contexts, aiming to refine its use and encourage widespread adoption in knowledge engineering and related fields.
Santos et al. (Fri,) studied this question.