Process Mining is a combination of data mining and algorithmic reverse process modelling. Process patterns are traced on the basis of the collected log data that transactions cause in ERP or other workflow business systems. These transactions are reverse engineered by process mining algorithms. They depict the real processes which took place at the time of the respective data logs. It is normal that most company processes have many variants that are not registered in the firm's business process models (BPM). Process mining illustrates process data in the business workflow context and thus allows to find root causes that only become visibel when looking at the sequence of events.
Digital Process Twin
This process minng analyses results in a digital process twin that mimicks the as-is process. The digital twin serves then as a model to deliver insights, as a basis to identifiy deviations againt target processes and as a playground to run scenarios. Often, data attached to the digital twin deliver new insights as they are illustrated in the timely sequence of the occurence in a business process.
Value-added by Process Mining
Many companies carry locked-in potential unquantified in their processes. Real processes are often the result of organizational behavior and discovery leads to value from ent-to-end diagnostic insight across organzational boundaries. For many organisations the illustration of data in the context of the firm's workflow is better digestable. It makes the data actionable. The transparency by the digital process twin becomes then a foundation upon tangible improvements can be achieved.
Challenges and Benefits for Business
Process mining may support a variety of business objectives in the area of performance, digitalization, complaince and automation. Linking company KPIs to process mining objectives is key for successful improvement initiatives. It is important to acknowledge that there is a direct relationship between existing data in the current system landscape and business targets and requirements. In other words if the data model is incomplete, some challenges cannot be tackeld by process mining. Furthermore, the data extraction can be challenging. On the other hand many published reference show that process mining can serve as a very valuable starting point of the digital enterprise transformation.
Purchase-to-pay (by Celonis)
The following short video demonstrates process mining for the purchase-to-pay process. Purchase-to-pay is a process that links the upstream value contribution of suppliers to the firm. The process connects the procurement of goods and services to the accounting of the supplier remuneration. Thus, it may serve as a starting point to use process mining for discovery.