A workspace is a key concept in Machine Learning Platform for AI (PAI). You can use workspaces to manage computing resources and user permissions in a centralized manner. Workspaces provide AI developers with development tools for collaborative operations during the entire workflow of AI development and allow the developers to manage AI computing assets. Before you can use the features of PAI, you must activate PAI and create the default workspace. This topic describes how to activate PAI and create the default workspace.


  1. Log on to the Machine Learning Platform for AI console.
  2. On the welcome page, click Activate and Create Default Workspace.
  3. In the Activate and Create Default Workspace panel, view the region and billing method of the default workspace.
    The region of the default workspace is the region that you select when you go to the welcome page. If you want to change the region, select another region from the drop-down list in the upper-left corner of the welcome page.
  4. Activate associated services and grant permissions to PAI.
    Activate associated services
    1. Specify whether to activate the required storage and training services free of charge. The services include pay-as-you-go Object Storage Service (OSS), pay-as-you-go MaxCompute, and DataWorks Basic Edition. You are not charged for the activation.
      If you have activated OSS, MaxCompute, and DataWorks, the system automatically selects the check boxes. You do not need to modify the settings.
    2. Click Authorize Now below RAM Role Authorization and create the required service-linked roles for PAI.
    3. After you read the terms of service, select the check box.
    4. Click Activate and Create Default Workspace to create the default workspace.


After you complete the preceding steps, the system performs the following operations:
  • Activates PAI and the other selected services, such as OSS, MaxCompute, and DataWorks, in the current region.

    You can log on to the console of each service and check whether the service is activated in the region.

  • Automatically creates the default workspace.

    Workspaces that are created in the PAI console are connected to workspaces that are created in the DataWorks console at the underlying layer. You can separately log on to the PAI console and the DataWorks console to check whether the default workspace is created.

  • Associates the default workspace with the pay-as-you-go MaxCompute resource group and the pay-as-you-go public resource group in Deep Learning Containers (DLC). The resource groups are used as the default computing resources.
  • Assigns the algorithm developer role to all the RAM users that belong to the Alibaba Cloud account. By default, the system assigns this role to the RAM users. After the default workspace is created, we recommend that you go to the details page of the default workspace and modify the roles of each workspace member. For more information, see Modify the role of a member.
    Note If MaxCompute resources are added when you activate PAI and create the default workspace, the system automatically assigns the MaxCompute Developer role to all RAM users that belong to the current Alibaba Cloud account.
After the default workspace is created, you can create more workspaces to meet your business requirements. For more information, see Create a workspace.