3.1.2 About Operator Tasks

Operator tasks control how a workflow moves from one step to another. Use operators to model branching, looping, joins, parallel execution, dynamic task selection, sub-workflows, and other orchestration patterns.

Operators do not usually perform business logic directly. Instead, they control the execution path of other tasks, similar to programming constructs such as if-else, switch-case, loops, and parallel blocks. For example, a switch operator can choose a path based on input data, a fork and join can run tasks in parallel and wait for them to finish, and a sub-workflow operator can run another workflow as part of a larger workflow.

MicroTx Workflows supports most of the operators available in open-source Conductor. It also includes the Agentic Planner operator, which uses an LLM-enabled planner to evaluate runtime context, prompt goals, tools, and available tasks, and then dynamically determine the next actions in a workflow. Use this operator to build adaptive agentic workflows that can plan and schedule follow-up steps based on runtime inputs.

For more information about standard operators, see Operators in open-source Conductor.