Enterprise Metadata Governance Architecture (EMGA): A Design Science Framework for AI-Ready Metadata Engineering and Governance

Enterprise Artificial Intelligence (AI) increasingly depends on trusted, standardized, and well-governed metadata to enable data discovery, semantic interoperability, Retrieval-Augmented Generation (RAG), and enterprise knowledge management. However, enterprise metadata remains fragmented across operational systems, cloud platforms, Software-as-a-Service (SaaS) applications, and governance repositories, limiting organizations' ability to establish consistent governance and support AI-driven initiatives. Existing metadata catalogs and governance platforms primarily emphasize metadata management rather than continuous metadata engineering throughout the enterprise metadata lifecycle.This paper proposes the Enterprise Metadata Governance Architecture (EMGA), a technology-independent Design Science Research (DSR) framework that unifies metadata acquisition, metadata transformation, governance, and metadata product generation within a continuous enterprise lifecycle. The framework produces governed metadata suitable for Artificial Intelligence (AI) applications, including semantic search, Retrieval-Augmented Generation (RAG), and enterprise knowledge services.The architecture was implemented and evaluated using Microsoft Fabric through the ingestion of heterogeneous metadata from relational databases and REST-based APIs. The implementation successfully demonstrated automated metadata acquisition, metadata standardization, governance publication, schema evolution detection, and AI-consumable metadata generation. The evaluation validated the engineering of nine metadata entities and 431 standardized metadata attributes, demonstrating the feasibility of the proposed framework in a modern enterprise cloud environment.The research contributes a reusable enterprise metadata framework that extends conventional metadata governance by integrating continuous metadata lifecycle with governance automation and AI readiness. The results demonstrate that EMGA provides a practical architectural foundation for organizations seeking to prepare trusted enterprise metadata for modern analytics and AI applications.