Skip to main content

Celonis Product Documentation

Knowledge Model

Introduction to Knowledge Models

A Knowledge Model (KM) is the central place to store and share business knowledge within a Studio app. As part of a Package, a KM stores reusable business entity definitions such as Records, KPIs, Filters and Variables.


You can think of a KM as the bridge between the technical Data Model layer and the Business Logic of a Studio App. Thus, a KM decouples data access and PQL calculations on the one hand and business and application logic on the other. This makes it possible to create Apps that are largely independent of the concrete data model instance. Data Models and Knowledge Models differ in the following ways:

Some Use Case Examples for the Benefits of Knowledge Models
  • A KM makes it possible to adapt a KPI (or Filter etc.) and its corresponding PQL formula in one central place so that it only has to be adapted once, regardless of where and how often the KPI is used in the package.

  • By using a KM, changes to the Data Model require only minimal adjustments in the KM, while Apps and Action Flows remain unaffected.

  • The extension mechanism of KMs makes it possible to define multiple custom definitions for different domains or organizational units without redundantly copying shared definitions.

  • The disabling mechanism gives you the opportunity to disable certain parts of your Knowledge Model instead of deleting them. This is useful when working in extensions and wanting to disable certain parts of the base KM.

Advantages and Features of a Knowledge Model

The centralized storage of business knowledge in Knowledge Models and the exposure of said Knowledge Models to other Studio services such as Action Flows provides various benefits:

  • Increased maintainability: Definitions are stored centrally in one place and all instances update automatically if a definition changes.

  • Increased reusability of knowledge: Any assets can reference definitions from a KM. Definitions can be reused across Packages, the App Store and within KMs.

  • Semantic enrichment and abstraction from data: Definitions in the KM are described in a more human-readable way and provide a more accessible abstraction of the underlying data.

  • Improved customizability: Definitions can be easily adapted and changed.

Auto-generated Knowledge Model Content

When connecting a KM to a DM the first time, the following objects will be auto-generated based on the underlying Data Model:

  • Records: For each table in the underlying DM:

    • Attributes: For each column in a table in the underlying DM

  • Event Logs:

  • KPIs:

    • COUNT_TABLE for every auto-generated Record

    • Avg Events Per Case Cel Ap Activities Activity En Bseg

    • Filtered Count

    • Number Of Process Variants Cel Ap Activities Activity En

    • Process Variants Cel Ap Activities Activity En

    • Ratio

    • Total Throughput Time In Days Cel Ap Activities Eventtime

Version Control

Changes done within Studio are not visible to your Business Users until they are published to the team. This allows you to thoroughly test new features and configurations. The Studio saves all published versions, so you are able to change the current version to a previously published version.

When publishing to the team, the Views and Analyses will be shown to Business Users on their Apps. However, Business Users will not be able to see Knowledge Models or Skills that you have created in the Studio.

You can publish a Package by clicking on the "Publish" button in the top right corner of the Studio. An overlay will appear summarizing which package assets have been modified since the last publication. You need to name the new package version you are about to publish ("New Version").

Base vs Extension Knowledge Models

Often you'll want to reuse parts of a KM or an entire KM. However, just copying and pasting the content leads to rework, and changes would not be reflected in the source. By introducing the Base and Extension mechanism, you will keep the connection to propagate changes throughout all connected Knowledge Models.

KM Extensions

Once you have made changes to an object in your Knowledge Model Extension, you are able to revert this object to its Base version using the Revert to Base functionality in the Knowledge Model Extension.

Getting started with Knowledge Models

How to add a Knowledge Model to your App

Studio enables the easy creation of new Knowledge Models.

  1. By hovering over a package or folder name, a "+" icon appears. Click the "+" icon and choose Knowledge Model.

  2. Enter a name for your KM. Choose if your KM should be a base or an extension.

    1. Choose base if your Knowledge Model should not extend another Knowledge Model.

    2. If you choose an extension, your KM will be linked to another already existing KM and implicitly contain the business knowledge entities of this "parent".

  3. Click Create.

  4. The KM auto-generates Records and common KPIs, which can immediately be referenced from Views.

Main Concepts of Knowledge Models
  1. Browse the KM objects.

  2. Search specific KM objects (Hint: Only the sub-objects are searchable).

  3. Get a detailed overview of current KM objects.


Additionally, it is possible to:

  • Toggle the auto-generated content.

  • Add new KM objects. This will open the corresponding Visual Editor (VE).

  • Switch to the YAML version of current KM (advanced users).

Deep Dive Into Listing Overview

The listing overview contains all objects that fall into a specific category of Knowledge Model object. From this overview, you are able to create new objects or edit existing objects. This Listing Overview also allows you to access and copy to clipboard the following value declarations for each KM object:

  • Name: Which shows the human readable name for the object.

  • ID: Which shows the unique identifier for the object in the Knowledge Model. This must be unique for the object within its category.

  • PQL Formula/Value/Skill: Shows the notation that defines this Knowledge Model object. The Main List for most objects shows a column with the PQL formula for each object, whereas the Variable Main Listing shows the Value for each Variable and the Action Main Listing shows the name of the Skill used to execute the Action.


You can also see the Source field for each of the objects. This field indicates the origin of the Knowledge Model object. This origin can be one of the following:

  • Autogenerated from the data model

  • Autogenerated from the data model but overwritten

  • Generated from a base knowledge model

  • Generated from a base knowledge model but overwritten

  • Custom added

Additionally, you can duplicate or delete a Knowledge Model object through the shortcuts available on hover within the Listing Overview: