Appearance
Client Onboarding Checklist
Use this checklist when deploying DMOps.ai for a new client. Complete every step before handing over to the client's admin.
Phase 1 — Infrastructure
- [ ] Agree on tenant ID (short, lowercase: e.g.
acme) - [ ] Agree on tier based on agent count and integration count
- [ ] Run
python acpfleet.py add <tenant_id> --tier <n> - [ ] Verify
GET /api/pingreturns{"status": "ok"} - [ ] Verify
GET /api/statsshows healthy dispatcher - [ ] Share the tenant URL (
<id>.app.dmops.ai) and initial admin credentials with client
Phase 2 — Configuration
- [ ] Set organisation name in
Settings → Organisation - [ ] Build the org chart — add client's human team members and AI agents
- [ ] Write or import skill documents for each agent role
- [ ] Connect required integrations and verify each with Test Connection
- [ ] Grant integration tool permissions to each agent
- [ ] Set budget limits per agent (monthly token and cost caps)
- [ ] Configure approval rules for any high-risk agent actions
Phase 3 — Agentic App (if applicable)
- [ ] Apply the app template:
python acpfleet.py apply-app <tenant_id> <app_name> - [ ] Review generated agents and skills — adjust system prompts for the client's context
- [ ] Connect client-specific integration credentials
- [ ] Run a test task end-to-end and verify it completes successfully
Phase 4 — User onboarding
- [ ] Invite the client's admin user (
Settings → Users → Invite) - [ ] Invite other human team members with appropriate roles
- [ ] Walk the client admin through the Admin Guide
- [ ] Walk end users through the End User Guide and mobile app install
- [ ] Run a live demo: create a task, have an agent run it, review the output in the inbox
Phase 5 — Handover
- [ ] Share links to the relevant docs sections with the client
- [ ] Confirm support contact channel (Slack / email)
- [ ] Schedule a 2-week check-in to review agent performance
- [ ] Document any client-specific customisations made during setup
TIP
Save a copy of this completed checklist in your internal project notes for each client. It's useful for debugging and for audits.