Artificial Intelligence For Root Cause Discovery In Distributed Platforms: Incident Evidence Mapping And Analysis

11 Jun
Posted on Author

Authors: Rebecca Hall, Jonathan Foster, Natalie Reed, Christopher Young, Chaitanya Srinivas, Rishi Kumar

Abstract: The increasing complexity of distributed platforms, cloud-native architectures, microservices ecosystems, and large-scale enterprise applications has significantly amplified the challenges associated with incident management and root cause identification. Traditional manual approaches to incident investigation often struggle to process the vast volumes of logs, metrics, traces, and operational events generated across interconnected systems, leading to prolonged resolution times and increased operational risks. This research explores the application of Artificial Intelligence (AI) for root cause discovery in distributed platforms through advanced incident evidence mapping and analytical techniques. The proposed framework leverages machine learning, graph-based dependency modeling, anomaly detection, natural language processing, and intelligent event correlation to automatically collect, organize, and analyze incident evidence from multiple heterogeneous sources. By constructing dynamic evidence maps that reveal relationships among system components, services, infrastructure resources, and operational events, the framework enables faster identification of causal factors contributing to system failures and performance degradation. The study further examines the role of predictive analytics and explainable AI in improving diagnostic transparency, reducing mean time to detection (MTTD), and minimizing mean time to resolution (MTTR). Experimental evaluations demonstrate that AI-driven evidence mapping significantly enhances incident investigation efficiency, improves diagnostic accuracy, and supports proactive operational decision-making in highly distributed environments. The findings highlight the potential of intelligent root cause analysis systems to strengthen platform reliability, operational resilience, and service availability while reducing the complexity of managing modern distributed computing infrastructures.

DOI: http://doi.org/10.5281/zenodo.20637297