- Release Notes
- Getting started
- Notifications
- Projects
- Datasets
- Data Labeling
- ML packages
- Out of the box packages
- Pipelines
- ML Skills
- ML Logs
- Document UnderstandingTM in AI Center
- AI Center API
- Licensing
- AI Solutions Templates
- How to
- Basic Troubleshooting Guide
Project-level access control
To better protect sensitive business data stored as training or evaluation datasets, you can set up access control based on your roles.
You can set up permissions for both groups or users for the following pages:
- At tenant level:
- Project: everyone can access their
project, but only users with
Read
permissions on project-level can view projects they are not part of. - Licenses: read-only.
- Users: roles can be managed from specific page.
- Roles: you can create roles for users.
- Project: everyone can access their
project, but only users with
- At project level:
- Dataset
- Pipelines
- Data Labeling
- ML Packages
- ML Skills
- ML Logs (read-only)
- Settings: user management at project level
The following roles and permissions are created by default for AI Center:
- Administrator: all permissions granted
- Tenant administrator:
- Project: create, read, update, and delete
- Licenses: read-only
- Users: create, read, update, and delete
- Roles: create, read, update, and delete
- Profiles: read-only
- Project creator:
- Project: create and update
- Profile: read-only
- Data scientist:
- Dataset: create, read, update, and delete
- Pipelines: create, read, update, and delete
- Data Labeling: create, read, update, and delete
- ML Packages: create, read, update, and delete
- ML Skills: create, read, update, and delete
- ML Logs: read-only
- Project administrator:
- Dataset: create, read, update, and delete
- Pipelines: create, read, update, and delete
- Data Labeling: create, read, update, and delete
- ML Packages: create, read, update, and delete
- ML Skills: create, read, update, and delete
- ML Logs: read-only
- Settings: create, read, update, and delete
- All-projects administrator:
- Project: read
- Dataset: create, read, update, and delete
- Pipelines: create, read, update, and delete
- Data Labeling: create, read, update, and delete
- ML Packages: create, read, update, and delete
- ML Skills: create, read, update, and delete
- ML Logs: read-only
- Settings: create, read, update, and delete
Note: The All-Projects administrator role is only available on tenants created starting with April 2024. Users with this role can access all existing projects and have full permissions within these projects. - Users manager:
- Settings: create, read, update, and delete
By default, group and role assignments are assigned when provisioning a tenant and when creating a new project.
The following default groups and roles are assigned when provisioning a tenant:
- The Administrator group is added to all AI Center tenants with the Administrator role.
- The Automation Developer group is added to all AI Center tenants with the Project Creator role.
- The Automation User group is added to all AI Center tenants with no role.
- The Everyone group is added with no role.
The following default groups and roles are assigned when creating a new project:
- The Project Creator is added to the project with the Project Administrator role.
- The Administrator group is added with the Users Manager role. Additionally, if the project creator decided not to restrict user access while creating the project, the Everyone group is added with the Data Scientist role.
The profile page displays two tables with permissions information on two levels:
- Tenant level
- Project level
The tenant table is static, only displaying permissions at tenant level.
In the project table you can view permissions at project level, choosing the project you are interested in from the drop-down list. You can also search for a project by typing in the drop-down field.