Google Cloud Launched Borderless Lakehouse Architecture
The platform now supports multi-vendor data integration to reduce the costs of enterprise AI agent operations.
Updated on Sept. 23, 2026 in Data Centers
Live Poll
Do you prefer cloud providers that allow you to use your data on any platform?
Google Cloud has released a borderless lakehouse architecture designed to unify data across disparate vendor environments and formats. This architecture, which is now available, uses cross-cloud interconnects to provide access to information without requiring full data migration.
Why it matters
This strategy targets the operational inefficiencies of fragmented data estates that drive up costs and complicate AI workflows. By streamlining access, Google Cloud intends to reduce token consumption and unnecessary tool calls for enterprise agents.
The architecture incorporates open table formats like Apache Iceberg to standardize data handling across environments. It leverages cross-cloud interconnects to enable native access, replacing the need for wholesale data relocation.
The players
Google Cloud
A division of Alphabet providing infrastructure, platform services, and AI solutions for enterprise data management.
SAP
A multinational software corporation specializing in enterprise resource planning systems.
Workday
A provider of enterprise cloud applications for finance and human resource management.
ServiceNow
A software company offering a cloud-based platform for workflow automation.
The details
The system utilizes a knowledge catalogue where AI agents enrich metadata to curate context more effectively. By automating metadata enrichment, the architecture reduces the frequency of unnecessary MCP (Model Context Protocol) tool calls—the standard used by AI models to interact with external data sources. This technical integration allows enterprise software such as SAP, Workday, and ServiceNow to operate within a unified lakehouse structure regardless of their underlying cloud vendor.
Timeline
September 23, 2026: Google Cloud officially introduced the borderless lakehouse and knowledge catalogue.
The Tech Race
This release marks a departure from proprietary data silos by adopting the Apache Iceberg open table format. It aligns Google Cloud with the broader industry movement toward interoperable data fabrics that minimize vendor lock-in.
Enterprises currently using SAP, Workday, or ServiceNow can now unify their data without the overhead of moving large datasets between clouds. Organizations should expect lower egress fees and optimized AI agent performance as these cross-cloud interconnects are implemented.
The takeaway
The move signals that the next phase of enterprise AI will be defined by cost-efficiency in data retrieval rather than just model size. Watch for the impact of this architecture on future monthly cloud infrastructure invoices.
Further reading
For more on evolving infrastructure trends, visit our Data Centers section.
Live Poll
Do you prefer cloud providers that allow you to use your data on any platform?
