Comparative Analysis of Low-Code Computation Systems

Abstract

The paper aims to systematically compare computation platforms where the development of custom computation applications is done visually. By this, we mean platforms equipped with a visual language to define the flow of actions or data, thus allowing us to treat them as low-code systems. The chosen platforms include two mature systems: Orange and Azure Machine Learning Studio, and also a newcomer -BalticLSC. For the purpose of the study, two sample computing tasks were created and executed on the three platforms. Based on this, the platforms were compared with each other taking into account the following characteristics: versatility, scalability, user entry barrier, cost of use, availability of documentation, maintainability and extensibility, availability, security, user interface friendliness, and variety of interfaces for input data.