This article was accurate at the time of publication in April 2026 and is reviewed regularly to reflect changes to the Workday platform and commercial model.
Everyone is talking about Workday AI. The conversations tend to start in the same place: which agents are available, what needs to be signed, and how quickly the efficiency gains will come.
What often gets skipped is a more fundamental question. Workday AI works by surfacing information from your tenant and acting on it, without a human sense-checking the output before it reaches your employees. That changes the stakes on data quality, configuration, and content in ways that are easy to underestimate.
Getting the foundations right before you activate is not AI preparation. It is the optimisation work that has always been worth doing. It just happens to make AI possible, and it determines how well AI performs once it is live.
The first place most organisations will feel the real-world impact of their AI readiness is the self-service agent. It is the first agent available today, and the first time AI will be answering your employees’ questions directly. There are five areas we recommend you have in good shape before you act on your Workday AI capability. The Foundations of Workday AI Readiness framework gives you guidance on what to consider at each stage.
The Foundations of Workday AI Readiness Framework
Knowledge Foundation
The agent answers employee questions by drawing on content that lives in your Workday tenant. If that content is not there, it has nothing to surface.
If you have Workday Help, your help articles need to be loaded and published, policy documents mapped to the right topic categories, and content kept current. If you use Guidance Workspace, it needs to be built out fully and linked to the right business processes, with content targeted at the right audiences.
Without Workday Help or Guidance Workspace, there will be limited information available for the Self-Service Agent to surface. An agent that cannot find an answer will tell your employees it cannot help. That is the fastest way to lose confidence in something before it has had a chance to prove its value.
Workday Help
Help articles loaded and published
Policy documents mapped
Content is current and reviewed
OR
Guidance Workspace
Guidance content complete
Linked to relevant business processes
Roles/ audiences correctly targeted
Business Process and Organisational Configuration
The agent operates within the structure of your tenant. If that structure is not accurate, it cannot work reliably.
Business processes need to be configured correctly and fully tested, with steps mapped to the right roles. Your organisational structure needs to reflect how the business works today. Supervisory orgs, manager hierarchies, job profiles, routing rules, approval chains, notifications. Most tenants have drifted in some of these areas since go-live. That is normal. It is worth knowing where you stand before AI starts operating within that structure.
Business Processes
Configured and fully tested
Steps mapped to the right roles
Completion path clear
Org Structure
Supervisory orgs accurate
Manager hierarchies correct
Job profiles maintained
Tasks and Inbox
Routing rules set up
Approval chains configured
Notifications reviewed
Security and Access
Before an agent starts acting on behalf of your employees, you need confidence that the right people can access the right things.
Worker, manager, and HR roles need to be clearly defined. Domain policies need to be correct. Employee self-service domains need to be activated, and help content needs to be accessible to the people who need it. Security models drift over time and are worth revisiting before anything new is introduced.
Security Groups
Worker, manager and HR roles are clearly defined
Domain policies correct
Self-service Permissions
Employee self-service domains activated
Help content accessible to employees
Worker Data Quality and Completeness
This is where most organisations have the most work to do, and where the consequences of getting it wrong are most visible.
With traditional Workday processes, a human is usually involved before an output reaches an employee. Someone checks, approves, or reviews. AI removes that layer. The answer goes directly to the person asking, based entirely on what is in your tenant. Which means errors in your data do not get caught before they land. They become the answer.
Core worker records need to be accurate and current. Absence plans need to be assigned correctly. Benefit plans need to be current. Pay group allocations need to be verified.
Absence tends to be where data quality issues are most common, particularly for organisations that have gone live in the last few years. If an employee asks how many days’ holiday they are entitled to and the absence plan assigned to them is wrong, the agent will give them the wrong answer. Across a workforce, that is a significant problem.
Core Worker Records
Contact details, locations and cost centres are correct
Job classification data accurate
Contractual information accurate
Plans and Entitlements
Absence plans assigned and correct
Benefit plans current
Pay group allocations verified
Governance and Ongoing Maintenance
Turning the agent on is not the end of the process.
Someone needs to own the content the agent draws on, with a defined review cycle and a clear process for identifying stale content. Usage needs to be tracked. Poor responses need to be flagged and acted on. Content gaps need to be filled. Without clear ownership, quality degrades. The agent will only ever be as good as the attention it receives after go-live.
Content Ownership
Named owners per topic area
Review cycle defined
Process in place for identifying and removing stale content
Post Go-Live Monitoring
Usage tracked
Poor responses flagged and acted on
Content gaps identified and filled
Note: This is not an exhaustive list, and the right starting point will vary depending on your organisation.
Getting all of this in place takes time. It is time well spent. The organisations that get the most from Workday AI are not the ones that move fastest. They are the ones that do the right work first.
We work with Workday customers to help them make sense of what Workday AI means for their organisation, where to start, and what is and is not worth doing right now. Get in touch with our team.