Generative Artificial Intelligence (AI), especially machine learning models that autonomously generate human-like content, has recently attracted significant attention in the education sector. This paper explores the potential of generative AI, including tools like ChatGPT, to shift from traditional outcome-oriented educational practices to a more balanced approach that values both the learning process and its outcomes. Traditionally, education has emphasized achieving predefined results, but the advent of generative AI tools, which enable students to easily produce tangible results, calls for a reevaluation of these practices. This shift suggests a need for a broader focus that encompasses the entire learning process leading to the final product, thereby promoting an educational practice that equally emphasizes both the journey and the destination of learning. Recognizing that the implementation of such practices, facilitated by generative AI, still requires exploration, this paper proposes a solution that integrates the experiential learning cycle and learning portfolio. This approach is designed to demonstrate the realization of process-outcome-balanced educational practices through the use of a pedagogical AI agent.