As enterprise AI evolves from simple prompt-response interactions to agentic execution, confusion often arises around what an "agent" actually is in a Workday ecosystem.
Industry discussions routinely conflate two fundamentally distinct architectural concepts: Workday Custom Agents (runtime agents executing within tenant boundaries) and Developer Agents (engineering copilots that help consultants build, test, and deploy Workday software).
For consultants, solution architects, and IT leadership, understanding the operational division between runtime agents and developer agents is key to planning successful Workday Extend and integration projects.
1. Defining the Two Agent Roles
| Dimension | Workday Custom Agents (Runtime) | Developer Agents (Copilots) |
|---|---|---|
| Primary User | Business users, managers, and enterprise employees | Workday consultants, developers, and solution architects |
| Execution Environment | Embedded inside the Workday tenant (Workday Illuminate / Extend) | Running locally on the developer's computer or workbench |
| Primary Mission | Answering tenant queries and triggering business processes | Authoring code, schemas, WQL queries, and packaging Extend apps |
| Data Scope | Tenant transactional data (workers, requisitions, payroll) | Development configurations, PMD/AMD models, and Studio artifacts |
| Lifecycle Stage | Post-deployment production runtime | Pre-deployment design, development, validation, and migration |
2. Workday Custom Agents: Operating Inside the Tenant
Workday Custom Agents are embedded assistants designed to streamline operational workflows for enterprise workers. Powered by Workday Illuminate and configured with custom tenant guardrails, these agents:
- Guide employees through complex open-enrollment choices.
- Assist hiring managers in generating tailored interview evaluations based on internal job profiles.
- Trigger business process transactions (such as expense approvals or leave requests) according to tenant security policies.
Because they operate at runtime, Custom Agents only know the schema, custom business objects (CBOs), and business processes that have already been created and deployed in the tenant. They do not build their own infrastructure or modify their own underlying application metadata.
3. Developer Agents: Building the Underlying System
A runtime agent cannot function without robust data models, secure business process definitions, and polished user interfaces. That is where Developer Agents enter the architecture.
A Developer Agent acts as an expert pair programmer and configuration assistant for the technical team. Instead of asking managers how many vacation days they have left, a developer agent assists consultants with:
- Data Modeling: Creating valid Custom Business Objects with proper keys, data types, and reference relationships.
- UI Authoring: Generating Presentation Page Models (PMD) with responsive layouts and accessible controls.
- Query Optimization: Authoring and testing Workday Query Language (WQL) queries to eliminate performance bottlenecks.
- Studio Integration Guidance: Debugging assembly steps, XSLT transformations, and REST web service invocations.
4. How kiweely Delivers the Developer Agent Experience
kiweely is engineered specifically as a developer copilot for the Workday ecosystem.
While enterprise clients implement Workday Custom Agents for their daily operational workforce, consulting teams use kiweely to construct those very capabilities faster and with far fewer errors:
- Context-Aware Pair Programming: kiweely reads your active project workspace, understanding the exact AMD schema, page model hierarchy, and integration targets you are developing.
- Local Privacy & Sandboxing: kiweely stores tenant connections and credentials locally on your computer with complete workspace isolation, ensuring client confidentiality across concurrent projects.
- Verified Deployments: kiweely interacts with Workday's real validation APIs, catching schema errors, invalid tag attributes, or malformed WQL statements before they ever reach a client tenant.
Both classes of agents have clear, complementary roles in modern enterprises: Custom Agents drive productivity for employees using Workday, while Developer Agents empower the consultants who build and maintain it.