Jupyter Notebooks have gained widespread popularity in research and education due to their ability to integrate code, explanatory text, and visualizations within a single document. This study examines their potential as a medium for preserving and transmitting domain-specific knowledge in machine learning research. A significant challenge encountered in research institutions pertains to the loss of specialized knowledge when projects reach their conclusion or researchers transition to new roles. Traditional documentation practices often fail to capture the depth of methodologies, design decisions, and experimental workflows, making it difficult for new researchers to build upon prior work. To address this issue, this study employs the ADDIE framework to design, develop, and evaluate Jupyter Notebooks as a knowledge transfer tool. Two notebooks were created with varying levels of explanatory detail. The first notebook incorporates comprehensive theoretical insights and contextual information. In addition, the second notebook was created with special attention to concise code documentation to support step-by-step execution and understanding. Eight researchers with machine learning backgrounds evaluated these notebooks through hands-on exploration, followed by semi-structured interviews. The results obtained indicate a strong preference for the detailed notebook, particularly regarding dataset descriptions, structured guidance, and visual support. However, concerns were raised regarding the effort required to create and maintain well-structured notebooks. The study suggests that integrating notebook creation into research workflows and leveraging automation tools may enhance sustainability. This study proposes Jupyter Notebooks as valuable supplement alongside more traditional documentation to improve research continuity and on-boarding processes, as they were never meant to replace formal documentation.