Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Towards the interoperability of low-code platforms

Iván Alfonso, Aaron Conrardy, Jordi Cabot

TL;DR

The paper tackles vendor lock-in in low-code platforms by evaluating current import/export limitations and proposing a model-based interoperability pipeline that uses a pivot metamodel (B-UML) within the BESSER framework. It couples formal text-to-model and image-to-model pathways to derive a B-UML intermediate representation, enabling migrations from a source LCP A to a target LCP B through M2T or spreadsheet-based imports. The approach is implemented as an open-source prototype demonstrating Mendix→PowerApps and PowerApps→Oracle Apex pipelines, with plans to extend to UI and behavioral models and to develop a user-friendly interface. This work advances cross-platform interoperability for LCPs, potentially reducing vendor lock-in and guiding future standardization and AI-enhanced migration methods.

Abstract

With the promise of accelerating software development, low-code platforms (LCPs) are becoming popular across various industries. Nevertheless, there are still barriers hindering their adoption. Among them, vendor lock-in is a major concern, especially considering the lack of interoperability between these platforms. Typically, after modeling an application in one LCP, migrating to another requires starting from scratch remodeling everything (the data model, the graphical user interface, workflows, etc.), in the new platform. To overcome this situation, this work proposes an approach to improve the interoperability of LCPs by (semi)automatically migrating models specified in one platform to another one. The concrete migration path depends on the capabilities of the source and target tools. We first analyze popular LCPs, characterize their import and export alternatives and define transformations between those data formats when available. This is then complemented with an LLM-based solution, where image recognition features of large language models are employed to migrate models based on a simple image export of the model at hand. The full pipelines are implemented on top of the BESSER modeling framework that acts as a pivot representation between the tools.

Towards the interoperability of low-code platforms

TL;DR

The paper tackles vendor lock-in in low-code platforms by evaluating current import/export limitations and proposing a model-based interoperability pipeline that uses a pivot metamodel (B-UML) within the BESSER framework. It couples formal text-to-model and image-to-model pathways to derive a B-UML intermediate representation, enabling migrations from a source LCP A to a target LCP B through M2T or spreadsheet-based imports. The approach is implemented as an open-source prototype demonstrating Mendix→PowerApps and PowerApps→Oracle Apex pipelines, with plans to extend to UI and behavioral models and to develop a user-friendly interface. This work advances cross-platform interoperability for LCPs, potentially reducing vendor lock-in and guiding future standardization and AI-enhanced migration methods.

Abstract

With the promise of accelerating software development, low-code platforms (LCPs) are becoming popular across various industries. Nevertheless, there are still barriers hindering their adoption. Among them, vendor lock-in is a major concern, especially considering the lack of interoperability between these platforms. Typically, after modeling an application in one LCP, migrating to another requires starting from scratch remodeling everything (the data model, the graphical user interface, workflows, etc.), in the new platform. To overcome this situation, this work proposes an approach to improve the interoperability of LCPs by (semi)automatically migrating models specified in one platform to another one. The concrete migration path depends on the capabilities of the source and target tools. We first analyze popular LCPs, characterize their import and export alternatives and define transformations between those data formats when available. This is then complemented with an LLM-based solution, where image recognition features of large language models are employed to migrate models based on a simple image export of the model at hand. The full pipelines are implemented on top of the BESSER modeling framework that acts as a pivot representation between the tools.

Paper Structure

This paper contains 18 sections, 6 figures, 1 table.

Table of Contents

  1. Introduction
  2. Evaluation of the interoperability among low-code tools
  3. Overview of our model-based migration approach
  4. Exporting the Model from LCP A
  5. Formal Export Method: Text to model
  6. Alternative Export Method: Image to model
  7. Generating and Importing the Model to LCP B
  8. Formal Import Method
  9. Alternative Import Method
  10. Tool support
  11. Discussion
  12. Lack of support for import/export models in LCPs
  13. Lack of standard interchange model
  14. Automatic model recovery techniques
  15. Incremental migration
  16. ...and 3 more sections

Figures (6)

  • Figure 1: Model-based migration solution for LCPs
  • Figure 2: Mapping of main concepts between Mendix and B-UML. Corresponding concepts are indicated by matching colors; for example, an Attribute in Mendix is mapped to a Property in B-UML
  • Figure 3: Overview of image to model pipeline with PowerApps data model
  • Figure 4: Library UML class diagram
  • Figure 5: From formal export to alternative import method
  • ...and 1 more figures