PQL

As a first example of Pull up function, I will show you First Time Right (FTR) that means process without rework activities (link). To determine whether each case is FTR or not, first I should determine rework activities, then judge if each case has rework activities or not. In this example I determined that rework activity name is started from Change. PQL to find such a string pattern is Change %, using wildcard % after Change. [Read More]

Pull up aggregation function, called PU function, like PU_COUNT, PU_SUM, PU_AVG, PU_MAX, PU_MIN looks similar to Standard aggregation function like COUNT, SUM, AVG, MAX, MIN. In Celonis, it is important to understand that output of PU function is not KPI but dimension. You can see Understand Difference between Dimension and KPI. If you know about SQL, you can imagine that PU function is similar to window function in SQL. As other functions, PU functions add dynamic column in grouping table. [Read More]

Process explorer in Celonis Analysis is default application to analyze process. First you can see main process pattern (called happy path), count of each step (activity) and count of path between two steps (connection). You can show minor activities and connections, or switch throughput time instead of count of connection. I think there are two ways to customize Process explorer. Change activity name Add more KPI against activity / connection Change activity name If you want to add detail information to each activity, you can change activity dimension. [Read More]

If you are beginner of creating Celonis Analysis, I believe you will first create OLAP table to understand how Celonis PQL works with data model. Points are, Dimensions are grouping key to calculate KPI, normally they are string (character) columns KPIs are calculated figures based on specific numeric columns, or count of rows of any columns, it must be numeric value in either cases Columns used as dimensions and KPIs are inplicitly joined based on data model Below screenshot is the quite simple example to show Dimension, KPI, and data model. [Read More]