DATA FABRIC ARCHITECTURE IN CLOUD EPM: SOLVING DATA SILO FRAGMENTATION FOR REAL-TIME CONSOLIDATION IN MULTINATIONAL ENTERPRISES

Multinational enterprises face critical challenges in achieving real-time financial consolidation due to data silo fragmentation across heterogeneous Enterprise Performance Management (EPM) systems. This research introduces a comprehensive Data Fabric Architecture specifically designed for cloud-based EPM environments, addressing fundamental limitations of traditional Extract-Transform-Load (ETL) approaches. Through empirical analysis across three Fortune 500 organizations, we demonstrate that the proposed architecture reduces data consolidation latency by 87.3% while improving data quality metrics by 94.2%. The framework integrates semantic layer abstraction, metadata-driven orchestration, and distributed query optimization to enable seamless cross-subsidiary data harmonization. Our findings reveal that organizations implementing the Data Fabric approach achieved real-time consolidation capabilities with 99.7% data accuracy, compared to 78.4% in traditional batch-oriented systems. The architecture supports dynamic schema evolution, automated lineage tracking, and intelligent caching mechanisms, fundamentally transforming how multinational enterprises manage financial data across geographic and organizational boundaries. This research contributes theoretical foundations for distributed EPM architectures and provides empirically validated implementation patterns for practitioners navigating digital transformation initiatives.