Modern enterprises increasingly depend on unified data platforms that integrate data ingestion, cloud computing, AI/ML pipelines, and business intelligence across heterogeneous environments. This paper analyzes multi-cloud data integration architectures spanning AWS, Azure, GCP, BigQuery, and Snowflake, evaluating their suitability for healthcare, FinTech, and retail analytics workloads across four dimensions: performance, governance, portability, and cost efficiency. We propose a layered reference architecture connecting ingestion, governed storage, analytics, and AI consumption; define practical evaluation metrics for cost efficiency, data freshness, and portability overhead; and compare platform ecosystems in terms of workload placement trade-offs. Our analysis demonstrates that no single platform dominates across all requirements. Instead, effective enterprise data platforms combine governed storage, reusable transformation layers, flexible ML services, and domain-specific consumption patterns. Optimal architecture depends on workload volatility, data residency constraints, team capabilities, and the balance between cross-cloud portability and platform-native optimization. These findings motivate a hybrid placement strategy governed by shared metadata, stable data contracts, and clear domain ownership rather than full-stack standardization on any single cloud provider.
Maruthavanan et al. (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: