Agentic workflow
A business process in which one or more AI agents carry out steps such as gathering information, deciding what to do next and calling tools, within boundaries defined by ordinary software. The workflow, not the model, defines what the agent may do and when a person must step in.
Why it matters
Agentic workflows can handle work that is too varied for fixed rules, such as reading unstructured requests, finding the right records and preparing an action. The risk is that open-ended autonomy makes behaviour hard to predict, test and audit.
Reliable agentic workflows keep the agent inside a defined process: explicit states, approved tools, validation of every action, limits on cost and retries, and human approval where the consequences warrant it.
In practice
For example, a UK property management company might use an agentic workflow for repair requests: the agent reads the tenant’s message and photos, checks the property record and contractor rota, drafts a work order and proposes an appointment, while a coordinator approves anything above a cost threshold.
Where Rodan fits
Rodan builds agentic workflows inside Governed Workflow Platforms, where state, permissions and approvals are handled by the platform. See also what is agentic AI and why does it matter for business.