Skip to content

Organizations overview

Alfe’s structure decides who can use which agents, and which accounts and knowledge those agents can reach. Set it up once, and new people and agents inherit the right access from where you place them.

You can work in one of two kinds of workspace:

  • Personal account — just you. Teams, projects, and Knowledge all work here.
  • Organization — a shared workspace with Members, Roles, and Settings in addition. Organizations also bill per seat on team plans.

Either way, all agents in a workspace share one credit pool.

Everything you share lives at a scope. Scopes nest from broad to specific:

  1. Organization — your whole workspace.
  2. Team — a group of people and agents. Teams can have sub-teams.
  3. Project — a piece of work. Projects can have sub-projects, and you can assign teams to them.
  4. Agent — a single agent.

A scope’s resources are available further down. A connection at team scope, for example, is usable by that team’s projects and agents. When a provider is connected at several levels, the most specific connection is the default.

Most resources can be set at more than one scope:

  • Connections — accounts your agents act through.
  • Integrations — tools and channels.
  • Files — shared files that sync to agents.
  • Knowledge — docs and a profile, at organization, team, and project scope.
  • Secrets — credentials agents use.

Each agent’s memory is the exception. It’s private to that agent, and it never flows through scopes.

If you can read identity data, the Identity section of the dashboard sidebar has three more views:

  • Contacts — every known person across your channels, with their linked platforms, tags, notes, and recent changes.
  • Organizations — identity-based access control groups.
  • Audit Log — tool execution and identity change history.

These hold the people your agents talk to, as distinct from the members of your Alfe workspace.