AI in mobile app development : enhancing UI/UX development with code generation
Ding, Chou-Ping (2025)
2025
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025051310969
https://urn.fi/URN:NBN:fi:amk-2025051310969
Tiivistelmä
The rapid growth of artificial intelligence (AI) tools has introduced new opportunities in frontend development, particularly through automated design-to-code solutions. This thesis investigates the practical feasibility of AI-assisted Figma-to-code tools in the context of mobile application development, with a specific focus on generating cross-platform React Native code. A hands-on testing approach was applied to Builder.io, a popular prototype-to-code tool, using two test scenarios: a simple static layout and a more complex interactive prototype.
Code generated by Builder.io was evaluated using five criteria: code quality, responsiveness, usability, reliability, and efficiency, and was tested within the Expo development environment. While the tool successfully translated structured Figma designs into modular and functional frontend code, several limitations were observed. These included styling inconsistencies, platform-specific behavior, and the inability to support multi-screen exports through a single workflow.
The findings suggest that while Builder.io and similar AI tools can significantly speed up the prototyping phase, they still require developer oversight for visual refinement, cross-platform validation, and implementation of custom logic. Overall, AI-assisted design-to-code tools show strong potential as workflow accelerators but are not yet fully capable of replacing manual development processes in professional mobile app projects.
Code generated by Builder.io was evaluated using five criteria: code quality, responsiveness, usability, reliability, and efficiency, and was tested within the Expo development environment. While the tool successfully translated structured Figma designs into modular and functional frontend code, several limitations were observed. These included styling inconsistencies, platform-specific behavior, and the inability to support multi-screen exports through a single workflow.
The findings suggest that while Builder.io and similar AI tools can significantly speed up the prototyping phase, they still require developer oversight for visual refinement, cross-platform validation, and implementation of custom logic. Overall, AI-assisted design-to-code tools show strong potential as workflow accelerators but are not yet fully capable of replacing manual development processes in professional mobile app projects.