From Lakehouse to Compliance Fabric: Building a Unified Cloud Data Ecosystem for Generative AI in Complex Corporate Tax Strategy

The paper provides an evaluation of how reliability, transparency, and reduce reproducibility of the corporate tax burdens can be increased by extending a lakehouse with the help of a compliance fabric, especially when the Generative AI models are utilized to interpret the rules. The paper focuses on comparing an improved system with a baseline architecture along the ingestion, lineage tracking, policy enforcement and reproducibility on the basis of quantitative metrics. The results have shown that there are major improvements in validated-record ratios, completeness in lineage, accuracy of policy options, and consistent Generative AI output. The working on the workloads was reduced since the metadata-based validation reduced the manual reconciliation. The multi-jurisdiction simulations also evolved into being stable and susceptible to tax regulations. The tax data operationalization with the help of the AI-driven compliance fabric is more defensible, traceable, and efficient