Healthcare providers continue to grapple with fragmented information technology (IT) systems that create data silos, inhibit unified access to patient records, and undermine care delivery, patient safety, and operational efficiency. This paper applies Enterprise Architecture (EA) principles and the Architecture Design Cycle to analyze the “as-is” state of healthcare IT infrastructure and to design a “to-be” solution built on FHIR-native microservices, an AI-powered Clinical Decision Support System (CDSS), and a hybrid cloud, hybrid database architecture. The proposed three-tier platform separates the presentation, application-logic, and data layers, and incorporates a defense-in-depth security architecture that reconciles interoperability with HIPAA compliance. Literature evidence indicates that microservices adoption can increase deployment frequency by up to 78% and reduce critical outages by up to 76%, while FHIR-native approaches can achieve interoperability success rates as high as 97%. Validated against real-world implementations such as Epic’s MyChart, Kaiser Permanente’s mobile application, and Cerner’s HL7-compliant integration suite, this paper demonstrates that EA-driven, FHIR-native microservices integrated with AI-powered CDSS offer a feasible pathway to overcoming healthcare IT fragmentation, enhancing patient safety, and reducing administrative costs.