Traditional vs. low-code development: comparing needed effort and system complexity in the NexusBRaNT experiment

Low-code platforms have gained attention in recent years as a visual and model-driven approach to application development. This approach uses drag-and-drop elements, model-driven logic, and graphical user interfaces to democratize development, making it possible for developers or business analysts of various skill levels to produce applications. In this experiment, we compare the traditional development approach with a low-code one using a platform that executes models in DEMO language, as an information and workflow system. We make this comparison with NexusBRaNT as its object, a system projected to manage information of a research project, which targets health professionals who perform cognitive rehabilitation. Our findings show that, in this particular case, the low-code approach could deliver a high-quality and functional application with an impressive reduction in needed time and implementation complexity. Namely, the standard approach took 888 hours of work, while using low-code, just 47.5 hours of work were needed, a 94.63% reduction in the needed effort. Regarding complexity, the traditional approach led to a total of 12310 lines of code implementing the system, while the low-code approach led to a total of 1723 database records, a reduction of 86%. This is highly relevant, especially for organizations that need to rapidly develop and deploy applications. The results of this experiment have important implications for the future of software development, particularly in the context of digital transformation and the increasing demand for agile and rapid application development.