Intelligent Ml-Based Workload Placement In Hybrid Clouds: Optimizing Cost And Sla In Modernized Systems

Abstract

The rapid adoption of hybrid cloud infrastructure has created unprecedented challenges in workload placement optimization, particularly in balancing cost efficiency with Service Level Agreement (SLA) compliance. This research investigates the application of machine learning techniques for intelligent workload placement in hybrid cloud environments, where organizations must dynamically allocate computational tasks across private and public cloud resources. Through a comprehensive analysis of workload characteristics, cost structures, and performance requirements, this study develops and evaluates an ML-based optimization framework capable of making real-time placement decisions. The research employs a mixed-methods approach, combining simulation-based experiments with case study analysis from three enterprise deployments. Results demonstrate that ML-based placement strategies achieve 34-42% cost reduction compared to traditional rule-based approaches while maintaining 97% SLA compliance. The study identifies key factors influencing placement decisions including workload predictability, resource availability, data locality, and security requirements. These findings contribute to the growing body of knowledge on cloud optimization and provide practical guidelines for organizations transitioning to hybrid cloud architectures.