Workday Data Cloud and Snowflake Zero-Copy Integration Goes GA

What the integration does

The integration makes Workday data directly accessible in Snowflake through a native inbound share via Apache Iceberg. According to Snowflake, Workday's semantic layer stays intact and governance is maintained at the source. The company says this brings people data together with the rest of an enterprise's data for analytics and AI, with money data coming soon.

Starting today, customers get bidirectional, zero-copy access to Workday's curated HR objects in Snowflake, including Worker, Position and Payroll data. Snowflake stresses that these are not raw database exports. They arrive with Workday's configurable semantic layer, so agents and analysts work with governed business concepts rather than raw column names. The integration is available on AWS and GCP.

Snowflake says the data can be combined with Cortex Agents, Snowflake CoWork and the Snowflake analytics layer, placing Workday's people and money context alongside Snowflake's AI and analytics infrastructure in one governed environment.

The problem with replication

Until now, getting Workday data into Snowflake meant building a replication pipeline, according to the post. That involved copying HR and financial data on a schedule, accepting a lag of 12 to 24 hours and maintaining the pipeline. Snowflake also notes that every replicated record of sensitive data sat outside Workday's governance model and widened the compliance exposure surface.

The company argues this matters for AI in particular. A workforce agent working from outdated headcount data, or a financial close agent using a stale general ledger, gives the wrong answer at the moment of decision, it says. With the zero-copy approach, Workday data is queried in place and never moved. Access is governed dynamically by org, title, reporting line and budget authority, and every action carries its audit trail, enforced at the source. Snowflake says customers do not need to build a separate governance framework, because the same controls used for native Snowflake tables apply.

How access works in both directions

To reach Workday data in Snowflake, customers create a zero-copy connector to their Workday tenant and mount the Workday Data Cloud catalog as a catalog-linked database. They can then query Workday Data Cloud tables from an existing Snowflake account without ETL pipelines. Two access paths are offered:

• Workday Data Lake: data materializes on a refresh schedule the customer defines, by choosing catalog objects and frequency. Snowflake describes it as suited to analytics, AI model training and reporting where near real-time freshness is not required.
• Workday Live Data Query: runs directly against the live Workday tenant over a standard SQL connection via JDBC or Python drivers. Data is current at the moment of the query, and the same access rules that apply inside Workday apply when querying from Snowflake. Snowflake positions it for operational dashboards and AI agents that need data measured in minutes.

The integration also works in reverse. Through Workday Prism Data Management, customers can connect Snowflake tables to Workday as external tables, zero-copy. They appear inside Workday as native data, joinable with Workday's HR and financial records, with no ETL and no duplication.

Snowflake gives an example: sales performance data in Snowflake could be joined with worker data and shown in a Workday discovery board. A Workday agent could then help managers identify which employees need support, recommend learning modules and assign them automatically.

Availability and what it means for customers

Workday customers whose tenant is on Public Cloud can use Workday Data Cloud with Snowflake today. Customers on Private Cloud are advised to speak with their Workday account teams about an expedited upgrade to Public Cloud.

For companies that run HR and finance on Workday and analytics on Snowflake, the change removes the need for scheduled copies and the separate governance that came with them. Money data is still listed as coming soon, so the finance side of the story is not yet available. Snowflake also invites customers to a session on Oct. 28 covering how the integration works and the use cases customers have put into production.