INFORMATION TECHNOLOGY FOR DELAY RISK ASSESSMENT IN SCRUM-BASED IT OUTSOURCING PROJECTS BASED ON DIGITAL COMMUNICATION PATTERNS AND TEAM PRODUCTIVITY METRICS

DOI: 10.31673/2412-4338.2025.048919

Abstract

An information technology has been developed to enable proactive detection of delay risks in Scrum-based IT outsourcing projects through the integration of digital communication patterns with team productivity metrics. The relevance of this research stems from the fact that 47% of distributed Agile teams experience significant delays, critically impacting the competitiveness of Ukraine's IT sector with its annual exports exceeding $7 billion. The core problem lies in the absence of integrated early risk detection methods that combine communication pattern analysis with traditional productivity metrics. A composite risk index is proposed, aggregating four metric domains: communication health, velocity stability, predictability, and team engagement. Domain weights were determined using the Analytic Hierarchy Process based on expert assessments from five professionals with 5+ years of Scrum project management experience, achieving a consistency ratio of CR = 0.044. A sliding-window algorithm adapted from the CUSUM method was implemented to identify degradation trends 2-3 sprints before critical events occur. Experimental validation on synthetic datasets confirmed model accuracy of 79% and robustness to data noise up to 15%. Expert validation demonstrated high interpretability of results (4.8/5), a significant advantage of the white-box approach. The technology is tailored for Ukrainian IT outsourcing companies and ensures a low implementation threshold through the use of standard tools (Jira, Slack, Git) without requiring additional infrastructure. The system architecture includes multi-level visualization for different stakeholder categories and supports real-time monitoring via WebSocket.

Keywords: Scrum, risk management, IT outsourcing, digital communication, composite index, predictive analytics.