Abstract

Agile management has become a central organizational strategy for coordinating software development teams in rapidly changing environments. However, the effectiveness of Agile practices is often evaluated subjectively, relying on qualitative assessments rather than data-driven evidence. This study investigates the managerial determinants of Agile team performance by applying Random Forest and XGBoost models to an AI-driven developer dataset. The analysis identifies the managerial factors that exert the strongest influence on task completion outcomes, focusing on resource allocation, risk assessment, project assignment, planning accuracy, and AI-enabled optimization. Results show that real-time resource prediction, AI optimization effectiveness, project overrun percentage, and risk assessment score are the most influential predictors of Agile performance, demonstrating that effective Agile Management depends on adaptive planning, proactive risk control, and strategic workload alignment. The study contributes to Agile literature by offering an empirical, model-based framework for evaluating managerial effectiveness and highlights the role of machine learning as a decision-support tool for enhancing Agile performance. These findings provide actionable insights for organizations seeking to strengthen Agile governance through predictive analytics and evidence-based management.

Keywords: Agile Management, Machine Learning, Random Forest, XGBoost, Managerial Predictors

How to Cite:

Siddique, Q., (2026) “Analyzing the Effectiveness of Agile Management Practices Through Random Forest and XGBoost-Based Performance Modeling”, Agile Management 1(2), 92-107. doi: https://doi.org/10.63913/am.v1i2.129