User Roles

How do you add and remove users on V7’s Darwin? We explore the what, how, and why of user roles.

Welcome to V7! In our Academy Series, we teach you how to build AI with V7’s Darwin, a powerful engine that empowers any business to develop state-of-the-art AI models. In this tutorial, we tackle User Roles, from how they work, to the powers that come with each type of role.

V7 provides five different User Roles: Team Owner, Admin, User, Workforce Manager, and Worker. The Team Owner is the user that first created the team within the platform. As a result, this user has the most permissions, from assigning roles to removing datasets to deleting teams entirely.

Following Team Owner is the Admin role. This user has all the permissions of a Team Owner, barring the ability to delete the team or change the Team Owner.

Next, we have the User. This role has a number of permissions, including the ability to add datasets and new users. They can also adjust the permissions of any User, Workforce Manager, and Worker.

The Workforce Manager’s permissions are restricted to the datasets that they are added to. Here, they can review, assign, and review tasks - in addition to inviting Workers.

Finally, Workers (otherwise known as Annotators) are your labeling workforce. They have all the permissions they need to complete their tasks but have restrictions beyond that to protect the integrity of your pipeline.

In this video, we explore the detail of how these User Roles work, where to find their permissions, how to assign them, and how to use each User Role to create a well-functioning development process. With this video, you’ll learn how to structure your project, and your people, to set you up for AI success.

Looking for a little bit of extra detail? Head to our User role documentation to see a comprehensive breakdown of the powers that come with each role.

This cog at the bottom left takes you to team settings. Here, you'll find the member tab on the left. Everyone in your team is listed here, from pending invites, to full team members, to annotators with limited access. Use this invite field to add more team members. Enter their email, and press the paper plane, or enter.

If you've made a mistake and invited the wrong Larry to your team, You can always delete this invite with the trash can button. They will still receive an invite email, but the link will be invalid. If an invite link has expired because too much time has passed, you can always delete an invite and resend it.

Beside each user, there is a dropdown to their role. They can be annotators. These have very limited visibility of a dataset. They cannot assign images to other users or view any other user in the team. Annotators can request and complete annotation or review work. But they can't really do much else.

Users are much more powerful. Those are the data scientists or data managers in your team. They can create datasets, add data, generate export versions, export data, they can invite other users, as well as annotators, and they can also annotate and review images. They cannot delete common threads started by admins, and they cannot invite other admins to the team.

But they can essentially do anything operationally relevant to the job of a data scientist. Admins are just like users who can delete anything they want, invite anyone they want, and make important billing decisions like change a credit card number. They are also the only user class who can make model training requests.

The last remaining user role is the team owner. This is the user that first created the team and they can assign this role to another admin. They're the only user who can delete the team and the main point of contact for any terms and conditions issues. Finally, to remove someone from your team, Just press the trash button next to their name.

Don't worry. Removing someone from your team will not delete their account. Any notation they made will stay, but become authorless and any task assigned to them will be unassigned so that someone else may pick it up.