Research indexJournal article
Challenges in Developing Artificial Intelligence-based Systems using Classic Software Development Life Cycle
Rahman
Journal of Mathematics and Computing Science9 Dec 2024Open access
NobleIDark:/48914/w1/20260524/546706A7DOI10.24191/jmcs.v10i2.4267
Challenges in Developing Artificial Intelligence-based Systems using Classic Software Development Life Cycle
Rahman · Journal of Mathematics and Computing Science · 2024
NobleIDark:/48914/w1/20260524/546706A7DOI10.24191/jmcs.v10i2.4267
Abstract
Artificial intelligence (AI) tools have become an essential part of modern software systems, driving automation, predictive analysis, and intelligent decision-making. However, the development of AI tools using traditional software engineering methodologies, specifically the Software Development Life Cycle (SDLC), presents unique challenges. This paper explores the intricacies and limitations of the classic SDLC in the context of AI-based system development, covering issues related to requirements gathering, design, implementation, testing, deployment, and maintenance. The analysis highlights the need for flexibility and iterative development models to meet the dynamic nature of AI projects. This paper summarises possible considerations that need to be taken care of since building AI-based systems requires a more flexible and iterative approach to accommodate the uncertainty, complexity, and dynamism in its own development.