Today, Workday is announcing Workday Data Cloud, a new architecture that opens the System of Record directly to AI agents—which means teams no longer have to extract data, duplicate it, or rebuild security.
For 30 years, enterprise software has forced a trade-off: to make data smart, teams had to extract it, losing its native security and business logic. Skilled professionals became human middleware, pulling data out to analyze it and logging back in to act. At Rising today, we’re ending that trade-off with the launch of Workday Data Cloud—a unified, open architecture that gives AI agents direct, secure access to live Workday data and business logic. Teams no longer have to extract data, duplicate it, or rebuild security.
And when organizations keep their data connected to its semantic truth in Workday, the business impact is massive:
Fitness Lifetime Group (FLG) is leveraging Workday Data Cloud’s zero-copy federation to power central enterprise analytics across Snowflake and Power BI. By replacing legacy extraction pipelines and orchestration jobs with direct SQL access, FLG unifies core workforce metrics, financials, and external Salesforce operational data into a single, real-time reporting model with zero data movement and total governance.
Werner Enterprises is scaling report-based integration, freeing its data team to focus on outcomes rather than infrastructure. "Workday is part of a three-legged stool in our enterprise, and Data Cloud extends its value by enabling report-based integration at scale,” said David Cavanaugh, manager, data solutions. “It feels like magic because we can focus on insights and outcomes instead of worrying about compute."
“It feels like magic because we can focus on insights and outcomes.”
David Cavanaugh
manager, data solutions
Werner Enterprises
In the agentic era, this shift matters more than ever: organizations can't take live operational action where data goes to rest—only where it actually lives. Workday processed 1.4 trillion transactions last fiscal year, 96.5% resolving in under a second, on a platform serving more than 80 million users. Today, that operational ground truth opens natively to agents.
We believe that the world’s most critical workflows can be run by autonomous AI, and the safety and trust of those workloads demand a unified architecture built explicitly for agents. That’s why Workday’s Data Cloud is built on three principles:
Open: Data should be universally accessible. Native zero-copy connectivity, powered by open standards (Apache Iceberg and Trino query federation), eliminates data duplication, vendor lock-in, and expensive Extract, Transform, and Load (ETL) pipelines.
Secure: Trust cannot stop at read-only access. Complete enterprise security powered by native contextual security and deterministic grounding manages what agents can do.
Headless: Your business logic is no longer trapped behind a user interface. Every routing step, approval chain, and semantic rule is exposed via Model Context Protocol (MCP) tools and skills natively for agents and agentic harnesses to execute against in real time.
We believe that the world’s most critical workflows can be run by autonomous AI, and the safety and trust of those workloads demand a unified architecture built explicitly for agents.
Foundation models are brilliant, but out of the box they have no idea how your business actually runs. The industry is bolting them onto the outside of the enterprise and forcing them to reverse-engineer business context from flat, stateless tables. An agent that has to guess your business rules will guess wrong.
To act safely, an agent needs three things the extract-first architecture can't provide: data it can trust, context it can inherit, and an identity it can be held to. As agents move from answering questions to taking actions, the question stops being "Can this agent see this record?" and becomes "Can this agent do this thing, with which tool, right now, and on whose authority?" It's not just the data businesses need to secure anymore. It's also the action.
As agents move from answering questions to taking actions, the question stops being "Can this agent see this record?" and becomes "Can this agent do this thing, with which tool, right now, on whose authority?"
Closing that gap required rethinking our architecture from the ground up. Workday is shifting from simply allowing data extraction to turning the world's most trusted system of record into the Trusted Agents Platform: the place agents go to get work done, not just get data.
That shift depends on how agents access data, context, and action at every layer of the stack. Workday Data Cloud handles this through four layers that work together:
Open Data Core brings operational and analytical data together in a single foundation. With rich hierarchical and temporal business data available directly in Workday Agent Lake, you have a live copy of your Workday data in open Iceberg format for large-scale analytics, alongside your third-party datasets.
Data Access and Processing delivers that data wherever teams already work, with Workday security and guardrails applied. Its key capability, Live Data Query, replaces slow extracts with sub-second, zero-ETL queries across live transactions and the Agent Lake. Zero-copy federation extends that access to the platforms teams already use.
Context Layer gives agents a governed understanding of what the data means, so a metric like turnover is defined the same way every time.
Agentic Experiences lets agents act on the data, not just query it. Agent-Ready Tools for Data give them governed ways to work with Workday data, and every call runs through the Workday Agent Gateway. Anything beyond an agent’s guardrails goes to a human.
By combining these four layers, Workday Data Cloud enables four distinct execution patterns in production:
Zero-Copy Lakehouse Federation: This enables data teams to connect Workday workforce and financial objects with external supply chain or CRM data directly within Snowflake, Databricks, or BigQuery without duplicating records.
Zero-ETL BI Bridge: Human analysts can now run real-time Power BI or Tableau dashboards directly against live operational state and historical trends using native JDBC connectors under User-As-Self security.
Deterministic Analytical Agents: First-party Workday Agents (such as Financial Test Suite or HR Analyst) and third-party tools query semantic models via MCP, retrieving 100% accurate, audit-ready calculations without schema guesswork.
Governed Operational Agents: Autonomous agents execute multi-step workflows, reallocating project headcount or adjusting procurement budgets, headlessly across any front door, bounded by real-time threshold guardrails and rollback protection.
Still at Rising? Catch the Workday Data Cloud Product Strategy & Vision session live and visit our express labs to see this architecture in action.
Not at Rising? Stream the session on-demand or explore the Workday Data Cloud User Guide to see how it opens the System of Record to AI agents.
More Reading
Developer Agent, Custom Agents, and Agent-Ready Tools are now generally available as part of Workday Extend AI—a comprehensive developer toolkit to build and run agents lawfully inside Workday’s guardrails.
At Workday Rising today, we're introducing new Sana capabilities that help employees get work done across HR, finance, and IT by coordinating the right agents, tools, and people behind the scenes. Watch a quick demo.
Workday's Talent Acquisition Agent handles routine hiring tasks automatically, keeping hiring moving 24/7. Recruiters spend less time on admin, so they can focus on engaging top talent and making better hiring decisions faster.
