CLOUD-NATIVE MACHINE LEARNING OPERATIONS FOR AUTOMATED HEALTHCARE CLAIMS ANALYTICS

Abstract

Healthcare organizations increasingly rely on cloud-native technologies to process large-scale insurance claims while ensuring scalability, security, and operational efficiency. Traditional claims processing systems face challenges in handling rapidly growing healthcare datasets, maintaining model performance, and supporting continuous deployment of machine learning solutions. Cloud-Native Machine Learning Operations (MLOps) combines containerization, orchestration, automation, and continuous monitoring to streamline the complete machine learning lifecycle for healthcare analytics. This paper proposes a cloud-native MLOps framework for automated healthcare claims analytics that integrates secure enterprise healthcare data, scalable Kubernetes-based infrastructure, automated model training, continuous deployment, and real-time monitoring. The proposed architecture enables intelligent claim prediction, fraud detection, denial prevention, reimbursement optimization, and explainable decision support while maintaining governance and regulatory compliance. Continuous feedback-driven model retraining improves prediction accuracy as healthcare policies evolve. The framework enhances operational efficiency, reduces manual intervention, accelerates claims adjudication, and supports intelligent healthcare decision-making through scalable cloud-native infrastructure and production-ready machine learning operations.