Model-Driven Engineering (MDE) leverages models to abstract complexity and automate software development, but its rigid workflows often hinder adoption in dynamic projects. Agile methodologies like Scrum offer flexibility, yet integrating them with MDE remains challenging due to manual modeling efforts. This paper proposes using Large Language Models (LLMs) to enhance MDE within Scrum, automating tasks like model generation and validation. We outline a workflow where LLMs act as intelligent assistants across Scrum sprints and present a case study of a ticket reservation system, showing reduced modeling time and improved agility. Challenges such as LLM accuracy and tool integration persist, but the approach offers a practical step toward merging MDE’s rigor with Agile adaptability.