Intel is a global leader in the semiconductor industry and one of a few remaining Integrated Device Manufacturers (IDMs). To support its vision of IDM 2.0, Intel is transforming the company into two primary business organizations, Intel Foundry and Intel Product, with a centralized corporate function. Each business organization will operate an independent enterprise analytics platform with strict data separation, a dedicated catalog (external Hive metastore), and a silo security model. With Databricks and Unity Catalog, Intel will unify multiple dedicated external Hive metastores into a centralized, secure and governed Unity data platform that enforces clear boundaries while preserving data ownership and enabling controlled data sharing across the enterprise.
Building a separation-first data strategy
Intel’s transformation program required the company to firmly separate its foundry, product, and corporate operations. The shift went beyond financial reporting and demanded a data architecture capable of reflecting that business structure.
“We needed to create separation from both a security and governance standpoint,” said Eric Messenger, Director of Data and Analytics IT. “And we needed the ability to audit and prove that clearly.”
Historically, systems and processes were blended across the enterprise. Intel moved from a single system to three: one for foundry operations, one for product operations, and one for corporate operations. The data platform had to enforce clear boundaries between domains while still enabling centralized insight across the enterprise.
Designing governance at enterprise scale
Rather than relying solely on policy controls within a single environment, Intel architected a domain-based governance model aligned to business boundaries. As part of this effort, the company migrated from external Hive metastores to Unity Catalog, adopting a multi-metastore architecture aligned to environments and domains.
Today, Intel operates more than 25 Unity Catalog metastores across development, QA, production, and disaster recovery environments, an approach designed to meet enterprise-scale governance and availability requirements. This structure enables each business segment to maintain centralized visibility within its domain, while helping to keep sensitive data appropriately separated.
Where shared reporting is required, Intel uses Delta Sharing within Unity Catalog to enable zero-copy data exchange across domains. This allows teams to collaborate without duplicating data, while maintaining a single source of truth and clear governance boundaries across domains between foundry and product domains.
At the same time, more than 38,000 tables in pre-production, test and production, as well as schemas, are today governed by a unified permissions and lineage model, supporting thousands of users and hundreds of developers globally. This overall architecture pushes enterprise limits while maintaining stability and high availability.
Strengthening governance with unified control
Migrating from an external Hive Metastore to Unity Catalog materially strengthened Intel’s data governance posture.
Intel’s revised operating model required governance controls to enforce strict separation between the foundry, product, and corporate domains, without fragmenting the data landscape. By adopting Unity Catalog as the centralized governance layer across its separate Product, Foundry, and Corporate metastores, Intel established unified security management, enabled row-level security where required, and introduced capabilities, such as end-to-end lineage across all data assets within a data lake, that deliver significant business value. This unified control plane, powered by Unity Catalog, preserves the deliberate isolation between business domains at the metastore level, while still enabling secure, auditable data sharing for enterprise analytics.
Teams can now define roles, assign permissions, and enforce fine-grained access policies from a single governance framework. Row-level and column-level security further restrict access to sensitive financial, customer, and operational data based on business function and user role, controls that are consistently applied and centrally managed across the Product, Foundry, and Corporate metastores.
Unity Catalog lineage provides additional transparency. “Unity Catalog delivers one of the most robust and intuitive lineage capabilities available today,” said Pradeep Baluvaneralu, Principal Engineer, Cloud Database Architecture, Intel IT. “It automatically captures deep, end‑to‑end visibility across tables, views, notebooks, workflows, Delta shares, and Genie spaces, making it easy to trace data movement and understand upstream and downstream dependencies. And because lineage is captured automatically with zero developer overhead, it scales effortlessly across our enterprise environments.”
“Running business-critical applications on Databricks meant that a robust disaster recovery strategy was non-negotiable,” said Sachin Arora, Data Platform Product Owner & Architect at Intel. “Together, Intel and Databricks designed and built a comprehensive BCDR solution on Unity Catalog, giving us the resilience and failover capabilities needed to keep critical data workloads running without interruption.”
These capabilities give Intel greater transparency, accelerating impact analysis, troubleshooting, compliance, and change management.
Powering AI on a governed foundation
With 2.8 petabytes (pre-production and production) of enterprise data unified and governed, Intel is accelerating advanced analytics and AI initiatives. The company leverages Databricks MLOps capabilities, including feature store and Model Serving, to operationalize machine learning across centralized datasets.
As Intel evaluates generative AI use cases, including natural language interfaces and multi-agent systems, governance remains foundational. Fine-grained access controls help to ensure that automated systems operate within defined security boundaries, accessing only the data aligned with user permissions.
By combining strict domain separation with unified analytics and AI, Intel has built a data platform that supports innovation while reinforcing enterprise-grade governance.
Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.
