Scalable and Extensible Cloud-based Low-Code Model Repository

Low-code development platforms (LCDPs) are becoming increasingly common in the software industry. By leveraging visual diagrams, dynamic graphical user interfaces, and declarative languages, these platforms support the development of full-fledged applications in the cloud. However, given the rapid evolution of these platforms, they encounter a fleet of challenges and limitations. Therefore, to address these issues, it is important to study them at their core conceptual level and their technologies. Model-Driven Engineering (MDE), along with cloud computing, forms the core foundation for these platforms. Although MDE has made significant progress, empirical studies show that practitioners still face challenges that hinder the broader adoption of MDE practices. The first obstacle is the need for efficient support for discovering and reuse of existing model artifacts. As a result, resources are wasted on developing similar tools and extensions, compromising the productivity and collaboration benefits of model-based processes. Second, modeling environments are deployed locally, leading to scalability, extensibility, collaboration, and performance issues. Consequently, modelers are compelled to download artifacts and executables to their local machines and then perform tedious configuration while installing MDE tools prior to utilization. Throughout this dissertation, we have attempted to advance the state of the art toward understanding and supporting cloud-based modeling in terms of LCDPs. Therefore, we have directly addressed issues of scalability and extensibility of modeling infrastructures and processes through the development of a cloud-based low-code model repository. This approach goes beyond the typical implementation of repositories with simple storage and query capabilities. We provide a large-scale repository and services for low-code engineering (LCE). The implemented repository enables access, persistence, discovery, and reuse of modeling artifacts via scalable and extensible approaches and infrastructures. In the LCE context, core services are containerized, orchestrated, and deployed as cloud services. The repository's functionalities can be extended via its remote API or by adding functionality in the form of extensions and services. Finally, an integrated web-based search platform and various domain-specific languages are devised to support various mechanisms for the composition, discovery, and reuse of persisted artifacts and model management services.