MULTI-AGENT GENERATIVE AI ARCHITECTURE FOR END-TO-END HEALTHCARE ADMINISTRATIVE AUTOMATION

Abstract

Healthcare administrative operations involve complex workflows including patient registration, eligibility verification, prior authorization, medical coding, claims adjudication, payment processing, denial management, provider communication, and regulatory compliance. Conventional administrative systems depend heavily on manual intervention, rule-based automation, and disconnected enterprise applications, resulting in delayed processing, increased operational costs, administrative errors, and inconsistent decision-making. Recent advancements in Multi-Agent Generative Artificial Intelligence (GenAI), Large Language Models (LLMs), RetrievalAugmented Generation (RAG), autonomous intelligent agents, and cloud-native computing enable collaborative healthcare workflow automation with explainable decision support. This paper proposes a Multi-Agent Generative AI architecture for end-to-end healthcare administrative automation by integrating autonomous AI agents, enterprise knowledge retrieval, machine learning, predictive analytics, and cloud-native orchestration. The proposed framework continuously coordinates specialized intelligent agents responsible for patient services, claims processing, coding validation, prior authorization, provider communication, compliance monitoring, and revenue cycle optimization. Experimental analysis demonstrates significant improvements in workflow efficiency, operational scalability, decision consistency, administrative productivity, and healthcare service quality compared with conventional administrative automation systems. Keywords: Multi-Agent AI, Generative AI, Healthcare Administration, Large Language Models, Retrieval-Augmented Generation, Intelligent Agents, Healthcare Claims, Revenue Cycle Management, Enterprise AI, CloudNative Computing.