Process Discovery

Process discovery is finding out how a business process actually runs, including its steps, variants, delays and exceptions, so it can be documented, improved or automated.

Common techniques are workshops and interviews with stakeholders, documentation reviews and BPMN modelling, process mining of system event logs, and task mining of desktop activity. Many programs combine them, because logs show what systems recorded while people describe what is missing.

Automated process discovery uses algorithms to build a process model from an event log, with a case identifier, an activity name and a timestamp per event. Process mining tools such as Celonis, Disco, Apromore and the open-source PM4Py do this.

Process mining is one family of process discovery techniques, along with conformance checking and performance analysis. Discovery in general also includes manual methods such as interviews and workshops. See the Process Mining Software page.

BPMN (Business Process Model and Notation) is a standard graphical notation for documenting processes. Discovered processes are often drawn in BPMN so business and technical teams share one model. See the BPM Software page.

Discovery identifies repetitive, rule-based steps that are good automation candidates before an RPA project starts. See the RPA Software and Task Mining pages.

Links were checked on 31 October 2026. Descriptions follow each source's own pages and are not rankings. Camunda Modeler's latest desktop release is from GitHub; the other release facts come from GitHub and PyPI.