Orvoq has three separate levels of role, and holding a role at one level says nothing about your access at another. This trips people up more than anything else in the product, so it's worth being precise about.
| Role | Can do |
|---|---|
| Owner | Everything an Admin can, plus billing, plan changes, deleting the org |
| Admin | Create workspaces, invite/remove org members, assign workspace roles |
| Member | Exists in the org directory only — no access to anything until assigned to a workspace |
| Role | Can do |
|---|---|
| Owner | Manage Knowledge, Files, Agents, Models, and policies; invite/remove members |
| Editor | Create sessions, use enabled agents and models |
| Viewer | Read-only access |
| Role | Can do |
|---|---|
| Owner | Invite/remove participants, manage AI participants, archive or delete the session |
| Editor | Converse, edit artifacts, create tasks, comment |
| Viewer | Read-only, can comment if allowed |
| Guest | Session-only, time-boxed — no workspace or org access implied |
Models and agents don't hold any of these roles. Instead, they're granted specific capabilities— read a document, search the web, send a message — visible on each agent's profile. There's no such thing as making an AI agent an “Owner.”
If someone's listed in your organization but greyed out with no workspace shown, that's not an error — it's someone who was added to the org directory without being assigned anywhere yet. Assign them to a workspace whenever it's actually needed.
Was this article helpful?