Beyond Data Integration: A Trust-Centric Architecture for Customer Identity Resolution, Master Data Governance, and Intelligent Enterprise Data Platforms
Year: 2026
Type: article
Abstract: Enterprise data platforms are commonly evaluated through measures such as connectivity, source onboarding, pipeline reliability, processing speed, and latency. Although these measures are important, they do not establish whether the resulting data can actually be trusted for business decisions. A technically successful integration may still produce duplicate customer identities, incorrect record associations, missing consent information, or attributes whose origins cannot be established. This paper proposes a trust-centric architecture in which trust is treated as a first-class property of enterprise data rather than an implicit consequence of integration and data-quality processes. The framework connects customer identity resolution, master data governance, data provenance, confidence assessment, stewardship, policy enforcement, and intelligent data consumption through a common trust substrate. Customer identity is treated as an evidence-based claim rather than an unquestioned fact, while attribute-level survivorship, source reliability, reversibility, and usage rights become explicit architectural properties. The paper further examines the implications of AI agents becoming direct consumers of master data and proposes trust-aware retrieval, external action gating, and closed feedback as essential controls. The resulting architecture moves beyond the traditional objective of making data available and instead focuses on making data explainable, traceable, appropriately governed, and fit for its intended use. The paper argues that an intelligent enterprise data platform should communicate not only what a customer record contains, but why each important value should be trusted, how it was established, and under what conditions it can safely support automated or human decisions.
Author: Somnath Banerjee
Language: English
Topic: Data Quality and Management
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7235429772
