RESEARCH ON THE IMPLEMENTATION OF GENERATIVE ARTIFICIAL INTELLIGENCE TO OVERCOME THE ISSUE OF RESOURCE AND TIME LOSS IN THE MODERN WEB APPLICATION DEVELOPMENT CYCLE

Abstract

In this article, a comprehensive analysis is presented of the modern web application development lifecycle, which typically combines Agile methodologies with DevOps practices, revealing the primary factors leading to resource and time overruns. The author points to ambiguous or frequently changing requirements, the accumulation of technical debt due to routine tasks and insufficient time for refactoring, inadequate test automation and challenges with reproducing defects, as well as communication difficulties in teams–especially distributed ones. Within this context, the article examines the role of generative artificial intelligence (GAI), capable of automating routine operations, writing or enhancing code, generating test cases and documentation, and in some cases providing design solutions.A review of scientific and practical publications shows that although Agile methodologies, DevOps practices, and continuous integration and delivery (CI/CD) tools receive considerable attention, the synergy of these approaches with the capabilities of generative AI has not been adequately explored. Notably, there is no complete overview of how AI can influence all phases of web development – from requirement gathering and alignment to product deployment and support. The author underscores the necessity of a systemic approach that encompasses both technological and organizational dimensions, also highlighting the risks linked to GAI usage (security vulnerabilities, potential “hallucinations,” and licensing issues).The central goal of this article is to evaluate the effectiveness of generative AI in addressing the “bottlenecks” in web development that lead to significant time losses. It is demonstrated that integrating GAI into Agile sprints can expedite backlog formation and the generation of both code and tests, while its inclusion in DevOps workflows simplifies the automation of CI/CD pipelines and infrastructure configuration. In addition, the article provides practical recommendations for thorough code review and the adaptation of project management methodologies to accommodate AI-generated content.The conclusion is that generative AI can substantially boost developer productivity, reduce repetitive tasks, and shorten release cycles when integrated with well-established Agile/DevOps processes. Future research should focus on real-world usage scenarios of GAI, the development of performance metrics (e.g., ROI), and the creation of guidelines for the safe and legally compliant use of language models. Given the rapid evolution of contemporary LLM solutions, establishing unified quality standards for generated code and standardized training for development teams remains highly relevant.Such a holistic approach will enable the full realization of GAI’s potential while minimizing security and regulatory risks.