A Survey of Natural Language-Based Editing of Low-Code Applications Using Large Language Models

In recent years, Large Language Models (LLMs) have showcased an impressive ability for natural language (NL) understanding, code generation, and logical reasoning. This provides the potential to significantly speed up development times by integrating this technology into software development workflows. Similarly, Low-Code Development Platforms (LCDPs) are already in use for reducing development effort and lowering the entry barrier to who can become a developer in the first place. This poses the question whether these technologies can be combined in order to enable end-users to edit an application via NL while experienced developers can still work on the same app using a regular LCDP and benefit from its advantages. To asses whether this proposal has been realised yet (and if so, to what extend), a literature survey is necessary. This paper presents such a survey, outlining how LLMs have been used to edit low-code applications, and especially Oracle Application Express (APEX) apps. It identifies an open research gap in this direction.