Measuring AI-Driven Developer Productivity in Agile Software Development Random Forest Regression Analysis of Performance Metrics and Tool Integration

The rapid integration of generative artificial intelligence (AI) tools in software development has created an urgent need to understand their impact on developer productivity and project outcomes. By examining AI-driven performance metrics for agile software development, this research addresses the current gap in empirical evidence on the effectiveness of generative AI in handling complex, real-world software projects. Traditional productivity metrics such as lines of code (LOC), feature completion rates, and defect density are inadequate for measuring productivity in AI-assisted environments. The study proposes new metrics including AI-assisted code generation, accuracy in defect detection, reduced developer effort, improved code quality, and reduced cognitive load to better assess the transformative potential of AI tools such as Git Hub Co-pilot and Open AI Codex. The research explores how these tools are reshaping traditional development workflows by automating routine tasks, enabling early bug detection, and improving project management processes. By analysing data from code repositories, issue trackers, and collaboration tools, the study provides evidence-based insights into developer productivity patterns. The findings show that AI-powered development tools can significantly reduce manual effort, accelerate software delivery, and improve overall quality, while allowing developers to focus on high-level design and strategic problem-solving, ultimately addressing the productivity paradox in modern software engineering.