Start, stop, and manage the lifecycle of your AI teams.

Deploying a Team

Each team runs in its own isolated environment. Here's how the lifecycle works.

Team Statuses

StatusMeaning
DraftTeam is not yet fully configured
StoppedConfigured but not running
CreatingEnvironment is being set up
RunningLive and processing messages
ErrorDeployment failed
ArchivedPermanently deactivated

Starting a Team

  1. Make sure the team has all required blocks configured
  2. Make sure your company has a positive balance
  3. Click Start on the team detail page
  4. Wait for status to change from "Creating" to "Running" (usually 30-60 seconds)

Your team is now live and ready to work.

Stopping a Team

  1. Click Stop on the team detail page
  2. The environment shuts down
  3. Infrastructure costs are calculated and deducted
  4. Agent memory is preserved for next start

Stopping a team does not delete agent memory. When you start it again, the agents can pick up where they left off.

Archiving

If you no longer need a team:

  1. Stop the team first (if running)
  2. Click Archive
  3. The team is hidden from your active view

Archived teams can be unarchived later if needed.

Costs While Running

While a team is running, you're billed for:

  • Infrastructure — $0.02/hour for the running environment
  • LLM usage — per-token costs based on the models used

Both are deducted from your company balance in real-time. See Usage Tracking for details.

If your company balance reaches $0, all running teams in that company will be stopped automatically.