As someone exploring data analytics as the next step in my career, the concept of Process Mining is an exciting discovery. It combines my new journey into data analysis with my previous experience in finance/business operations.
I explore the concept of Process Mining in a two part article series. Here is the second medium article if you want to understand the details of *how* process mining is done
Process Mining is a methodology to use data from business applications (e.g. ERP and CRM) for discovering, monitoring and improving business processes. Celonis is one of the pioneers in Process Mining.
What is process mining² ?
- Information Systems create event logs as a business activity completes: an order is received, a product is delivered, a payment is made. The logs make visible how the work is happening, who did it, how long it takes etc.
- An event log consists of multiple cases. Each case has a unique ID assigned to it and consists of a set of events. These events are equivalent to the activities undertaken in a process. And each of these activities has a set of attributes. Example: A purchase requisition for material ‘ABC’ with its unique ID ‘1234’ is created by a business unit. It will go through a series of ‘events’ like creating a purchase order, creating an invoice, scanning an invoice, payment and delivery. These are the series of events that ID 1234 went through with a clear start and end time.
- The process mining technology captures the digital footprints from any number of systems throughout an organization. It organizes them in a way that shows each step of the journey to complete that process, mines it for business insights and highlights with any deviations from the expected path.
Why is process mining useful?
Process Mining is helpful in a myriad ways for a business process: discovery, conformance, process improvement.
A) Process discovery uncovers the workflow of a business process avoiding the lengthy whiteboarding and interviewing of stakeholders. It builds a visualization of the process and demonstrates every step in the cycle. The discovery happens by extracting process models from an event log
B) Process Conformance indicates how much variation exists from the identified ‘as is’ process. Celonis provides analytics with built in KPI like throughput time. You can also create custom built KPIs using Process Query Language (Celonis uses a variation of SQL called PQL). Automated Conformance Checking, based on Process Mining AI, evaluates current process performance against the reference process model and instantly highlights areas that fail to conform. Process conformance reveals inconsistencies and can help with process standardization where possible.
C) Process Improvement: The ‘Action Engine’¹ module in the Celonis Intelligent Business Cloud. It operationalizes the process insights in order to create intelligent predictions and recommendations as well as to initiate activities using machine learning. Think of Netflix/Amazon recommendations based on your viewing/shopping history. The first step of the Action Engine is to autonomously analyze data across processes and information systems. Next it relates the detected improvement opportunities to users — in a timely and personalized way. Finally, it proposes action to a predefined user/user-group or executes this action in the source system (e.g. by triggering a bot or starting a workflow).
The Action engine lets one set triggers by initiating signals to the users, routing rules to define the recipient (Head of Procurement) and assignee of the task (procurement manager). There are 3 ways to set triggers:
a)Rule based triggers: Using PQL(Celonis’s query language) one can set rules for the process mining data. The system will constantly analyze the data vis-à-vis the rule and forwards the resulting recommendation to the end user.
For example in a scenario where late checks of payment blocks for external invoices in paper form. These lead to a loss of cash discounts. A rule is set to check such payment blocks earlier. The rule can be defined via PQL in the same way as an Event-Condition-Action formula. If the set of rules is in force, the end user receives a notification (by email, Slack or any other communication channel) in the case of an external invoice (as a condition and trigger of the rule) to check the payment block on time.
b)Triggers by classification of comparable process: Machine learning is used to classify processes into groups with similar characteristics for instance by country, material group. Based on the analysis of the processes within a comparison group, predictions about the further course of the process and recommendations for current process flows are made.
For example take a scenario involving placing a purchase order. Based on similar orders in the past, a supplier can be recommended from whom similar/same articles were ordered. This reduces the research effort of the buyer.
c)Triggers by predictive process monitoring: Through predictive analysis, potential future problems can be detected and preventive actions can be taken in order to avoid unexpected situation, e.g., processing delay or late invoice payments. To prevent this, a rule is created and a recommendation for action is initiated. In addition, the next activity of a process can be predicted, including the time it will take place, for example a delivery to the customer. Below is an example of a visual of predictive capabilities.
For anyone wanting to learn more about the fundamentals of how to get started with process mining. Check out my second medium article here
Sources
[1]Badakhshan, Bernhart, Geyer-Klingeberg, Nakladal, Schenk, Vogelgesang The Action Engine — Turning Process Insights into Action. http://ceur-ws.org/Vol-2374/paper8.pdf
[2]Davenport and Spanyi (2019).What Process Mining Is, and Why Companies Should Do It. https://hbr.org/2019/04/what-process-mining-is-and-why-companies-should-do-it
[3]www.celonis.com
[4]Online Training available for Analyst and Data Engineer for the Celonis Intelligent Business Cloud at https://www.celonis.com/training/
[5]http://www.heeh.nl/en/processmining
[6]Santoso, A., Felderer, M. Specification-driven predictive business process monitoring. Softw Syst Model 19, 1307–1343 (2020). https://doi.org/10.1007/s10270-019-00761-w
[7]https://www.celonis.com/process-mining/process-mining-white-paper#how-process-mining-works
