Sarcouncil Journal of Engineering and Computer Sciences

Abstract: The exponential growth of cloud-native applications and microservices architectures has introduced unprecedented complexity in IT operations management. Traditional mon-itoring approaches are insufficient to handle the dynamic, distributed nature of modern systems. This paper presents a comprehensive observability and AIOps (Artificial Intelli-gence for IT Operations) framework that integrates machine learning, real-time analytics, and intelligent automation to enhance system reliability, performance optimization, and incident response. The proposed framework combines three core components: (1) com-prehensive data collection from metrics, logs, and traces, (2) AI-powered anomaly detec-tion and root cause analysis, and (3) automated remediation and predictive maintenance. Through experimental evaluation on production-like environments, the framework demon-strates significant improvements in mean time to detection (MTTD) by 68%, mean time to resolution (MTTR) by 54%, and overall system availability by 12%. The results indi-cate that AIOps-driven observability can substantially improve operational efficiency while reducing manual intervention and operational costs.