Skip to main content

Celonis Product Documentation

Getting started with the ML workbench

Description

You have your process data connected to the EMS and a data model ready to be analyzed.

In this tutorial, we show you how you can use the ML Workbench to access, process and push data from and to the EMS.

1. How to create a new ML Workbench

  1. In the EMS Header, click Machine Learning.

  2. Click Create ML Workbench.

    20873420.png
  3. A model is shown. Enter a Name for your Workbench.

    30998735.png
  4. An application key will be created automatically with the name [Name of Your Workbench (Workbench)].

    20873421.png
  5. Give this application key permissions (e.g. All permissions to the Analysis you want to work with, and the Data Model and Data Pool containing the data of the Analysis). For more details on how to manage permissions, watch the video below.

     
  6. After initializing, you are sent to the Jupyter Launcher. From here you can create a new Notebook. To learn more about Jupyter, see Development Environment (Jupyterlab) Documentation.

  7. To learn more about connecting to the other applications, pulling and pushing data, etc., check out the Celonis Basics page.

2. How to use the getting started tutorials for the ML Workbench

You can access a getting started guide directly inside the Workbench by following these steps:

  1. In the Launcher, create a new Python 3 Notebook.

  2. In the first cell, enter:

    from pycelonis import notebooks
  3. The File browser on the left will now be populated with new files. Double-click on any of those to get started on a certain topic.