Appearance
Workflow: Deploy → Configure → Scale
The official AgentBrain rollout pattern. Three phases. Each phase combines one or more SOPs.
Phase Diagram
flowchart LR D[Deploy] --> C[Configure] C --> S[Scale] S -.feedback.-> C
Phase 1: Deploy
Install AgentBrain on your infrastructure: cloud, on-premise, or air-gapped.
SOPs in this phase:
- On-Premise Deployment (or pick Cloud SaaS for the fastest path)
Exit criteria:
- Health check returns ok
- UI loads at your hostname
- TLS is valid
Phase 2: Configure
Connect your knowledge sources, set up RBAC, and define agent workflows.
SOPs in this phase:
- RBAC Setup
- Knowledge Vault Setup
- Channel Deployment (per channel)
Exit criteria:
- SSO works for at least one non-admin user
- At least one knowledge collection is fully synced
- At least one agent is invokable on at least one channel
- Audit log shows the expected events
Phase 3: Scale
Roll out agents department by department with full audit and control.
Steps in this phase:
- Pick the next department to onboard
- Define the agents that department needs
- Grant the department's roles access to those agents
- Connect the knowledge collections the department needs
- Bind agents to the channels the department uses
- Train the department in a 30-minute session
- Monitor adoption and audit log activity for the first week
- Iterate based on questions agents fail to answer well
Department Rollout Order
Pick the department most likely to validate adoption (often Customer Support or Engineering). Once that succeeds, expand outward.
Feedback Loop
Phase 3 surfaces gaps that flow back to Phase 2:
- Missing knowledge collections → run Knowledge Vault Setup again
- Permission mismatches → tune RBAC
- Channel friction → add or reconfigure channels
Related
- Pricing for tier requirements
- SOPs
- Case Studies for production deployment patterns