Databricks can now query Workday HR and finance data in place, without copying it

Databricks released a beta connector on October 8 that lets companies query their Workday HR and finance data from Databricks without building pipelines or keeping a second copy. The Workday Data Connect federation connector reads the Apache Iceberg tables that Workday shares through Workday Data Cloud, straight from Workday's managed storage.

Databricks says it is one of the first Workday Data Cloud launch partners to offer this kind of connector publicly.

How it works

1. A Workday administrator shares approved HR and finance tables through Workday Data Cloud and gives read access to an integration system user.
2. A Databricks administrator creates an OAuth connection and a foreign catalog in Unity Catalog.
3. Unity Catalog reads the table metadata, and Databricks compute reads the data from Workday's object storage.
4. The administrator sets table-level permissions, and users query the data with SQL or Databricks' Genie assistant.

Access is read-only, so Workday stays the system of record. Databricks handles the OAuth token exchange with Workday's Iceberg REST catalog.

Jonathan Rogers, chief information officer of FactSet, said the connector provides "a governed, zero-copy way to discover and catalog Workday data in Unity Catalog."

Requirements

The connector needs a workspace with Unity Catalog enabled and Databricks Runtime 19 or above. A workspace admin turns it on from the Previews page, and Databricks says customers can contact their account team to join the beta. It is separate from the existing Lakeflow Connect connectors that ingest Workday data into Databricks.

Why zero-copy matters for HR data

HR and payroll data is some of the most sensitive data a company holds. Every copy made for analytics is another place where access rules can drift from the source. Reading the data in place, under Workday's sharing rules and Unity Catalog permissions, cuts down on the copies.