The Process Intelligence, using the data registered by (or in) the information systems during organization's day to day operations, creates an objective knowledge of its real way of working.

According to the available data in the logs, complete models of workflows, organizational context and rules can be easily "learned". In other words, the required knowledge of the "as-is" situation is obtained, obliged starting  point of most activities focused on automating, improving or simply  understanding own operational behaviour.

Because no any a-priori model or piece of information is required, it can be applied in completely unknown situation, with no idea about what can be expected. But it also proves equally usefull for refining or completing partial or incomplete information.

Furthermore, Process Intelligence allows to evaluate the extracted models or to perform detailed benchmarks.

Different methodologies are used, according to the investigated perspectives (activity flow, organization or business rules) and the expected pieces of information (for example: process diagram, patterns, sequence clustering or classification, roles and relationships, decision tree, etc.). Some of them are derived from machine learning, social network and data mining fields, others are ad hoc alghoritms (some of them often collected under the extending process mining umbrella). All implemented in well known programming languages (mostly R, the envinronment for statistical computing and graphics) or analysis tools (Weka, Pajek).