Process mining with artificial intelligence

Celonis has won Hana Innovation Awards at Sapphire for the last two years in a row - in 2015 and 2016. What have you been working on since then?

Bastian Nominacher: Celonis' process mining technology is now very mature and offers corporations and mid-sized companies a powerful platform for analyzing their process data.

What we're looking at now at Celonis are technological innovations that enable companies to use process mining intelligently. Our latest achievement is Celonis Pi, a smart extension of our solution that has artificial intelligence.

In our offices in New York, Munich and Den Bosch in the Netherlands, our employees are continuously working - often together with local universities - on ever new functionalities, thus paving the way for innovations; especially in the field of machine learning and artificial intelligence.

Celonis is part of the SAP Startup Focus program. In the transition from newcomer to solution provider for Fortune 500 companies, how has SAP's support changed?

Nominacher: In fact, we have been part of the SAP Startup Focus program for a very long time and have grown a lot as a company during that time. The collaboration with SAP was quite crucial for us to further develop our process mining technology for SAP customers and to win more customers - and it still is.

With SAP as a partner, we can ideally align our sales and service to the needs of SAP customers. As the number of customers has naturally grown, so has our cooperation: We are now even an SAP Platinum Partner.

Collectively, we serve customers worldwide in all business areas. This is also reflected in our collaboration with analysts and our joint presence at events or trade shows - the best example being the upcoming Sapphire conference in Orlando.

Many companies still get deeper insights into key metrics via classic BI solutions. What are the differences to Celonis Pi?

Nominacher: Basically, process mining and classic BI solutions differ from each other in three points: data basis, goal of the analysis and visualization of the results.

Users of BI solutions are concerned with identifying individual aspects in their data via aggregated KPIs - they obtain meaningful results above all if their question is precise and their analyses are based on a large volume of data.

Process mining, on the other hand, does not start at the data level but at the process level and visualizes the implicit process knowledge already contained in the data.

Accordingly, process mining makes use of a collection of data in which the individual process steps are stored - information on time stamp, name and case key is already sufficient here.

The most important thing is that the view of the processes is open-ended. Celonis Pi analyzes processes and data not only in real time, but also in a forward-looking manner - and provides the causes of problems and measures for their solution at the same time.

SolMan 7.2 is the established solution for business process documentation. Is there a way to link the analyses from Celonis with the documentation from SolMan?

Nominacher: SolMan 7.2. places a great deal of focus on project and solution documentation. All created process steps are stored in a process library.

The solution thus offers ideal starting points for process mining. All the more so because it offers the possibility of creating BPMN-compliant process models - i.e. target process models - via the new process editor.

There are two different approaches for integrating the two solutions: On the one hand, it is possible to update the documentation based on the actual processes identified by process mining. Celonis reconstructs the actual processes from the raw data and exports them as a BPMN model - they are then loaded into SolMan.

On the other hand, Celonis offers the possibility to compare the target processes stored in SolMan with the actual state through its conformance checking. For this purpose, the target process is loaded from SolMan into Celonis. A comparison then shows how well the documentation corresponds to the actual processes.

How did you present at Sapphire Now 2017 and what are you planning next?

Nominacher: For us, Sapphire Now is a forum to show what is possible with Celonis. Also this year, we presented our solution and showed concrete application possibilities using the example of customers such as Akzo Nobel or Vodafone.

For example, users have reported how they optimize business processes and increase efficiency with Celonis. Compared to last year, we took a big step towards artificial intelligence and added machine learning capabilities to Celonis Process Mining.

The result is our intelligent extension Celonis Pi. The focus of our trade show appearance was to present the new functionalities and to make Machine Learning in the area of Process Mining live experienceable for visitors.

I am firmly convinced that it will only be possible to derive the highest added value from process analysis in the future with artificial intelligence. At Celonis, we will support customers with the continuous development of our process mining technology.