LLM4Reuse

Low-code development platforms afford fast and easy process automation. Their intuitive drag-and-drop interfaces enable employees without formal programming skills to participate in process automa-tion. However, these so-called citizen developers often struggle to create high-quality automation and face challenges in understanding, maintaining, or reusing existing low-code automation due to a lack of formal training. These issues can severely hamper the effectiveness of citizen developer projects or even prohibit potential citizen developers from partaking. In response, this paper presents an artefact aimed at lowering barriers for citizen developers and assisting them in their process automation. Using design science research, we create design principles and develop an artefact that illustrates the ap-proach using the example of robotic process automation. Building on a large language model with a custom front-end, the proposed artefact lowers barriers to citizen development by automatically inter-preting, documenting, and adapting complex low-code automation using natural language. We demon-strate the functionality of our solution and evaluate its value with practitioners.