Process Modeling and Automation
Nodes, flows, and variables
A process model is a directed graph of nodes connected by flows. Node families include start/end events, user input tasks, service tasks (script, integration, data write), gateways (XOR, AND, OR), subprocesses, and timer events. Process variables carry state; process properties are metadata. Process parameters pass data in at start.
Execution semantics
An XOR gateway takes exactly one outgoing path based on conditions evaluated in order. An AND gateway takes all paths. A subprocess runs a separate process instance and can be synchronous or asynchronous. Escalations and timer events handle timeouts. Error handling is configured per node with exception flows or boundary events.
Traps to watch for
- Assuming XOR evaluates all conditions. It evaluates in order and stops at the first true path.
- Confusing a subprocess with a process model called via a start process node - the latter is a separate instance with its own lifecycle.
- Forgetting that a user input task requires assignment (user, group, or expression) or the instance stalls.
- Believing a script task can write to a database without a data store entity or integration.
- Overlooking that process variables are not automatically persisted to records.
Rule of thumb
Trace a single token through the model. For every node ask: what data does it need, what does it produce, and who or what executes it? If any of the three is unanswered, the model has a gap - and exam distractors usually exploit exactly that gap.
Worked example
A process hangs after an approval. The user input task assigns to a group with no members. The instance waits indefinitely. The fix is to assign to a group with active members or to a dynamic expression that resolves to a user.