XREP: The XQuery Rule Evaluation Pattern—A Metadata-Driven Approach to Runtime Business Rule Processing in Native XML Systems

Modern SaaS systems increasingly rely on dynamic policy and rule evaluation to enforce tenant-specific business logic. Traditional rule engines introduce architectural complexity by requiring separate rule languages and execution environments distinct from the application and data layers. This paper introduces XREP (XQuery Rule Evaluation Pattern), a metadata-driven architectural pattern for runtime rule processing in native XML systems. XREP enables business rules to be stored as XML metadata and evaluated dynamically using XQuery, eliminating the rule–data impedance mismatch common in conventional rule engines. The paper formalizes the pattern, describes its implementation in a multi-tenant SaaS platform, and compares it with existing rule and policy systems such as Drools, XACML, Open Policy Agent, and Cedar. The results demonstrate how linguistic homogeneity between rules, queries, and data simplifies system architecture while enabling flexible runtime extensibility. The XREP pattern illustrates how declarative rule processing can be integrated directly within XML-native platforms to support scalable, metadata-driven application design.