Add Members
Invite team members and configure their access to datasets and training projects
Add members
Invite team members to collaborate on your organization's datasets and training projects. Configure access levels and project permissions when sending invitations.
PrerequisitesBefore adding members, ensure you have:
- Admin or Owner role in your organization
- Available collaborator slots in your subscription plan (see Resource Usage)
- Email addresses of the people you want to invite
Add a team member
-
Click the Settings icon in the bottom left corner of your workspace dashboard
-
Go to the organization settings menu (top navigation bar) and select your organization name
-
Click Members in the left sidebar
-
Click + Add Member in the top right
-
In the Invite Members dialog, configure the new member's access:
Field Description Options Email Member's email address Valid email address Organization Role Organization-wide permissions Admin, Member Dataset Which datasets they can access All Datasets, or select specific datasets Dataset Role Permissions level for datasets Owner, Editor, Viewer Training Which training projects they can access All Trainings, or select specific projects Training Role Permissions level for training projects Owner, Editor, Viewer -
(Optional) Click + Add Another Member to invite multiple people at once with different access settings
-
Check the Collaborators counter at the bottom (e.g., "Collaborators: 1/50") to ensure you have available slots
-
Click Invite Users to send the invitations
The invited members will receive an email with instructions to join your organization. If they don't have a Datature account, they'll be prompted to create one.
Invitations sentOnce you click Invite Users, invitation emails are sent immediately. New members can accept the invitation and access assigned resources right away.
Understanding roles
Organization roles
Organization roles determine overall permissions within your organization:
| Role | Can do | Typical use |
|---|---|---|
| Admin | Manage members, access all resources, create projects, manage settings (cannot manage billing) | Team leads, project managers |
| Member | Access only assigned datasets and training projects | Individual contributors, annotators |
Owner roleThe Owner role has full control including billing access. Owners are set during organization creation. To transfer ownership to another member, contact Datature support at [email protected].
Dataset roles
Dataset roles control what members can do with assigned datasets:
| Role | Permissions | Use for |
|---|---|---|
| Owner | Full control: upload, annotate, delete images, manage dataset settings, and delete the dataset | Dataset managers |
| Editor | Upload and annotate images, edit annotations, manage image-level settings | Annotators, data preparation team |
| Viewer | View images and annotations only, no modifications | Quality reviewers, stakeholders |
Training project roles
Training project roles control what members can do with assigned training projects:
| Role | Permissions | Use for |
|---|---|---|
| Owner | Full control: create workflows, start runs, modify settings, deploy models, and delete projects | Project leads, ML engineers |
| Editor | Create workflows, start training runs, view results | Model trainers |
| Viewer | View workflows, runs, metrics, and results only | Stakeholders, auditors |
Granting access to specific projects
You can grant members access to specific datasets and training projects rather than all resources:
Select specific datasets
In the Dataset dropdown:
- Click the dropdown (it defaults to "All Datasets")
- Select individual datasets by name
- Selected datasets will be shown in the dropdown
Members will only access the datasets you select.
Select specific training projects
In the Training dropdown:
- Click the dropdown (it defaults to "All Trainings")
- Select individual training projects by name
- Selected projects will be shown in the dropdown
Members will only access the training projects you select.
Mixing access levelsYou can grant access to All Datasets but only specific training projects (or vice versa). Configure dataset and training access independently based on your team's needs.
Inviting multiple members at once
To invite several people with different access levels:
-
Fill in the first member's details (email, roles, access)
-
Click + Add Another Member
-
A new row appears below with separate fields
-
Configure the second member's email and access settings
-
Repeat for additional members
-
Click Invite Users to send all invitations at once
Each member row can have different organization roles, dataset access, and training access. This is useful for onboarding entire teams efficiently.
Common scenarios
Adding an annotator
For team members who will annotate images but not train models:
- Organization Role: Member
- Dataset: Select the specific datasets they'll annotate
- Dataset Role: Editor (allows annotation)
- Training: Leave unselected or choose "None"
- Training Role: Not needed if no training access
This gives them access to annotate specific datasets without accessing training projects.
Adding a project manager
For team leads who need to manage all projects:
- Organization Role: Admin
- Dataset: All Datasets
- Dataset Role: Owner
- Training: All Trainings
- Training Role: Owner
This provides full project management capabilities without billing access.
Adding a read-only stakeholder
For stakeholders who need to monitor progress without making changes:
- Organization Role: Member
- Dataset: Select specific datasets they need to see
- Dataset Role: Viewer
- Training: Select specific training projects
- Training Role: Viewer
They can view progress and results without modifying anything.
Adding a contractor for one project
For temporary team members working on a specific project:
- Organization Role: Member
- Dataset: Select only the dataset for their project
- Dataset Role: Editor (if they need to annotate) or Viewer
- Training: Select only the relevant training project
- Training Role: Viewer or Editor based on needs
Limit access to only what's necessary, then remove them when the project ends.
Adding an ML engineer
For team members who will train and deploy models:
- Organization Role: Admin or Member (based on leadership role)
- Dataset: All Datasets or specific datasets
- Dataset Role: Owner or Editor
- Training: All Trainings or specific projects
- Training Role: Owner
This provides full control over model training and deployment.
Best practices
Start with minimal access
Principle of least privilege:
- Begin with Member organization role and Viewer project roles
- Grant access only to specific projects they need
- Upgrade permissions when their responsibilities grow
- Avoid "All Datasets" or "All Trainings" unless truly needed
This minimizes security risks while maintaining flexibility.
Document role conventions
Maintain clear role guidelines:
- Define when to use Admin vs. Member for your team
- Document which project roles apply to different job functions
- Share these guidelines with new members during onboarding
- Review and update conventions as your team grows
Clear conventions prevent confusion and access creep.
Batch invitations for teams
Efficient onboarding:
- Use + Add Another Member to invite entire teams at once
- Prepare email addresses and desired access levels beforehand
- Group members with similar roles to streamline configuration
- Send all invitations in one batch to maintain consistency
Saves time and ensures consistent access setup.
Verify collaborator limits
Check available slots:
- Review the Collaborators counter before inviting (e.g., "Collaborators: 1/50")
- See Resource Usage for your plan's limits
- Upgrade your plan if you need more collaborator slots
- Remove inactive members to free up slots
Ensures invitations succeed without hitting plan limits.
Common questions
What happens after I send an invitation?
After clicking Invite Users:
- Each member receives an email invitation to join your organization
- If they have a Datature account, they can accept immediately
- If they don't have an account, they'll be prompted to create one first
- Once they accept, they gain access to assigned datasets and training projects
Invitations don't expire, but you can remove the member to revoke the invitation.
Can I change a member's role after inviting them?
Not directly. To modify roles or access:
- Remove the member
- Re-invite them with updated roles and access settings
Their work (annotations, training runs) remains intact—only their access changes.
Do invited members count against my collaborator limit?
Yes. Each invited member (regardless of whether they've accepted) counts toward your plan's collaborator limit. See Resource Usage for your current usage and limits.
To free up slots, remove members who haven't accepted invitations or are no longer active.
Can I invite someone with just an email address?
Yes. You can invite anyone with a valid email address, even if they don't have a Datature account yet. They'll be prompted to create an account when accepting the invitation.
What if I don't see the dataset or training project I want to assign?
If a dataset or training project doesn't appear in the dropdowns:
- Make sure the dataset or project exists in your organization
- Verify you have Admin or Owner permissions to grant access
- Try refreshing the page if you just created the resource
If issues persist, contact [email protected].
Can members invite other members?
Only Admin and Owner roles can invite new members. Members cannot invite others or modify access settings.
If a team member needs to invite others, upgrade them to Admin role.
Related resources
- View member details — Check project access for existing members
- Remove members — Revoke access when members leave
- Resource Usage — Monitor collaborator limits and upgrade plans
- Team Settings overview — Complete team management guide
Need help?
We're here to support your VLMOps journey. Reach out through any of these channels:
Updated about 1 month ago