Celonis Product Documentation

We've renamed Primary Keys to Identifiers, with all existing Primary Keys automatically migrated to Identifiers. Identifiers are still optional and only need to be set to enable Delta Data Model loads and other functionalities that require identifying unique rows when building apps in Studio.

You can set or edit the Identifier for your tables from your data model diagram:

For more information, see: Executing delta extractions.

We've added filters to the Objects page and the Events page in the Objects and Events environment, so you can browse a list containing only the object types or event types you want to work with. You can filter the lists by selecting one or more of these tags:

You can pick more than one tag from a category, and we'll show object types and event types matching any of them. If you apply filters from more than one category, we'll show the object types and event types that match a selected tag in all of those categories (for example, Celonis object types in the Accounts Payable or Accounts Receivable process).

The filters you set stay in place when you move from the Objects page to the Events page and back, and when you reload the page. You can bookmark a filtered view or share it with other team members by copying the browser URL.

The object-centric data model explains filtering and other ways to explore your objects, events, and relationships.

The comment component allows Studio app viewers to add text-based comments within a view. These comments are now associated with individual rows in a Data Model table, so you must provide an identifier definition for the respective table in the Data Model before configuring your comments component.

Once the first comment is written in the component, an augmentation table is created and added to the Data Model automatically. This table contains information around the comment such as the author, the message, or the creation time. The table storing comment information gets updated in real time and can be queried in the PQL editor.

As an example, all comments created for one row of the BKPF table can be counted via PQL:

As comments are saved in augmented tables in the Data Model, they can also be added as columns in a table and used in linked views:

For more information, see: Comments.

You can now add one or more new attributes to an object type or event type while you’re creating the transformation for it. You can use this capability if you’re working with the SQL transformation and identify an attribute that would be useful to have, but didn’t get added in modeling. You can add attributes to a custom object type, a custom event type, or a Celonis object type.

Select the Add attributes button in the preview pane in the SQL editor to open a new window. Click Add to add each new attribute, name it, and pick a data type. When you save your new attributes, we’ll update the object type or event type, and warn you if other transformations for it besides the one you’re editing might be affected by your changes. For the instructions to work with the SQL editor, see Creating custom transformations.

The process for setting up value tracking in Transformation Hub has been changed to allow you to see how the selected filters impact the KPI values. The set up process has been changed from three steps to two by combining the existing "Filter Data" and "Set up Formula" steps.

In this new version, you select a date column and filters on the same screen as where you select KPIs. So when you choose KPIs, Transformation Hub generates a preview of how your filter selections impact the KPI values. Previously, the effects of your filters on the KPIs were not shown which led to incorrect selections and no value being tracked.

For more information on the complete setup process, see Setting up value tracking.

We've changed the way the SQL analyzer validates SQL statements that include the COALESCE() function. To ensure consistent behavior, all the operands for COALESCE should be of the same data type. If they aren't, COALESCE might return the result as a data type that you didn’t expect, causing downstream errors that are harder to troubleshoot.

Previously, our SQL analyzer allowed mixed data types for COALESCE. From this week, it’ll return an error if you try to publish your object-centric data model with a statement like that in a custom transformation. The error message looks like this:

Cannot infer return type for COALESCE; operand types: [....]

Transformations that you already published won’t stop working, but you’ll see the errors when you next attempt to publish. In order to publish the transformations, you'll need to resolve these errors by changing each affected SQL statement to use the same data type for all the operands for COALESCE.

For example, this statement mixes a column with the TIMESTAMP data type, and a timestamp represented using the STRING data type:

COALESCE(timestamp_column, ‘2024-04-16’)

You can fix this by giving the literal the correct data type:

COALESCE(timestamp_column, timestamp‘2024-04-16’)

For the instructions to work with custom transformations, see Creating custom transformations.

In Transformation Hub, you can now edit the framed value or the framed value link for a live opportunity. If you’ve started tracking the realized value, but you change your mind about the prospective value based on actual performance, resourcing changes, or a new insight, you can update the framed value to reflect your new expectations. For the feature documentation, see Transformation Hub.

We've been busy enhancing the Studio, your workbench for analyzing data, exploring processes, and building solutions. From May 2024, our enhanced Studio features will be in general availability for everyone to use. These streamlined interfaces make building views and solutions faster, more intuitive, and a much more joyful end-to-end experience for you.

The exciting enhancements to Studio include:

Our enhanced Studio experience will be available in May. We'll announce the exact release date two weeks prior to go-live, giving you time to be prepared.

For more information: If you want to want to know more about these changes:

When you create custom object types and event types, you can now use Japanese characters in the name of an object or event, and also in the names of its attributes. For the custom modeling instructions, see Creating custom object types and custom event types.

For this version of the JDBC Extractor, we’ve fixed some issues with the recommended metadata change handling option while configuring delta extractions.

For the instructions to update, see Installing and updating the on-premise JDBC extractor.

In the new Studio experience, you create augmented attributes in the Knowledge Model by directly referencing tables in the Data Model rather than referencing records. With a dedicated Migration Tool, you can turn your current record-based augmented attributes into a format which is ready to be used in the new Studio experience. For more information, see Migrating augmented attributes.

As part of the Notifications section in Admin and Settings we have added a new feature that allow the admin of the team to create “in-product” notification banners. Once these notifications are created and published they become visible to all their users.

This feature provides greater flexibility for admins when they want to communicate updates to their users. The updates might include:

We have also included a notification dashboard where admins can manage their published notifications.