Title: Vibe-Driven Model-Based Engineering

Abstract:

The escalating complexity and rising demand for modern software systems have created an urgent requirement for more effective development methodologies and tools. Contemporary challenges, such as the emergence of novel user interface paradigms, the integration of intelligent components, and the imperative for sustainability, necessitate new strategies for handling these issues. While model-driven engineering (MDE)—encompassing its recent evolution into low- and no-code development—has significantly enhanced software productivity and quality in recent years, the task of specifying and managing models is becoming increasingly difficult. Conversely, "vibe coding," a method gaining traction through its use of Large Language Models (LLMs) to convert natural language prompts into executable code, introduces risks related to code vulnerabilities, scalability limitations, and maintainability. Although some predict that vibe coding will supplant model-based engineering, this paper contends that the two methodologies are complementary rather than mutually exclusive. They offer distinct development pathways suited to different software types, scenarios, and user demographics. To this end, we propose vibe-driven model-based engineering, a novel framework designed to synthesize the strengths of both AI and MDE to hasten the creation of reliable, complex systems. We delineate the fundamental concepts of this integrated approach and discuss the future opportunities and unresolved challenges it presents for the software development landscape.

Source: arXiv Generated at: 2026-06-02 00:00:00 UTC