A MODEL FOR SELECTING ARTIFICIAL INTELLIGENCE TOOLS TO SUPPORT SOFTWARE DEVELOPMENT PROCESSES

• Published in: Vìsnik NTU <<HPÌ>>. Serìâ: Strategìčne upravlìnnâ, upravlìnnâ portfelâmi, programami ta proektami (Online) no. 2(9); pp. 45 - 49
• Main Authors: Kopp, Andrii, Nesterenko, Ivan
• Format: Journal Article
• Language: English
• Published: 17.03.2025
• ISSN: 2311-4738; 2413-3000
• DOI: 10.20998/2413-3000.2024.9.6

Integrating artificial intelligence (AI) tools into software development projects significantly improves the efficiency of various tasks within the software development lifecycle (SDLC). AI-driven tools embedded in integrated development environments (IDEs) improve developer productivity and code quality, and facilitate better interaction between project participants and AI-based systems. The main research directions for integrating AI into software development processes include adapting user interfaces for specific tasks, increasing trust in AI-based systems, and improving code readability. AI enhances several SDLC stages, including automated code generation, code review and defect prediction. Implementing AI tools in IDEs accelerates development, improves code quality and reduces defects. Machine learning and natural language processing play a critical role in improving software quality through requirements classification and defect prediction. AI-based solutions, such as recommendation systems and chatbots, support various software development processes, including requirements gathering. Therefore, a relevant scientific and practical challenge is to create a model for the justified selection of AI tools to support software development processes in order to improve project efficiency. This study proposes a mathematical model that minimizes the cost of using AI tools, while ensuring compliance with minimum requirements that affect project efficiency. The optimization model takes into account criteria such as pricing, integration, support and functionality capabilities, using normalized evaluations based on Gartner Peer Insights and other open sources. The objective function minimizes the total cost of AI tools, subject to constraints that ensure minimum acceptable evaluation scores. The developed approach enables a systematic selection of AI tools, thus improving the efficiency of software development projects.