Development of a Methodology for Automated Backend Code Optimization Triggered from a Frontend Web Application Using Generative Artificial Intelligence Models

In this paper, the authors introduce an innovative methodology for automating backend code optimization, triggered from the frontend of a web application and powered by advanced generative AI models. The proposed system integrates the Vue.js and Laravel frameworks, together with a PHP (Hypertext preprocessor) Parser and the OpenAI API (Application programming interface), enabling dynamic, real-time optimization without necessitating system downtime. In contrast to conventional approaches that rely on offline environments or CI/CD (Continuous integration and continuous deployment) pipelines, this methodology facilitates continuous performance enhancement during active usage of web application. An experimental evaluation involving 20 functions within a Laravel application demonstrated a 70% optimization success rate, a 62.6% reduction in execution time and an 18.18% decrease in code lines. This approach preserves semantic consistency and provides reports to developers for further implementation. By reducing errors and enabling ongoing refinements, the work makes a significant contribution to the field of web application development.