Skip to content
P

Platform admin quickstart

Provision organisations, approve signups, and monitor a PIE deployment as a platform admin.

This tutorial is for a platform admin — PIE staff who operate the platform for its tenants. It covers provisioning organisations, handling self-service signups, and knowing where to watch the running system. It assumes the stack is already deployed (see the operator handbook for that).

The platform admin role

platform_admin has cross-tenant read access and manages organisations and billing. It is the only role that can act across tenant boundaries; everything a producer or buyer does is scoped to their own organisation. Use it deliberately — cross-tenant reads are audited.

1. Provision an organisation

Every organisation — producer or buyer — enters PIE through one bootstrapper. From the Admin Console, create an organisation with a name, side (PIP or PIC), and plan tier. This creates the org, its first admin user, and the default settings (enabled locales, entitlements). The first admin receives an invitation token to complete registration; hand it to them for the Producer quickstart or Buyer quickstart.

2. Handle self-service signups

If public signup is enabled, prospective organisations register themselves and land in an approval queue. Review each request, then approve or reject it. Approval provisions the organisation exactly as if you had created it by hand, and the requester completes registration. Signup mode is controlled by deployment configuration — see the environment reference.

3. Manage plans and billing

Each organisation has a plan tier that maps to a set of entitlements (rate limits, feature access, usage caps). Review tiers and per-tenant subscriptions from the console. See Plans & limits for what each tier grants, and the Plans & billing concept for how metering works.

4. Monitor the running platform

Health and metrics are served on an internal ops port that is never exposed publicly — Prometheus scrapes it on the private network. Grafana provides dashboards and email alerting behind authentication. The tracing pipeline is opt-in. For the full picture — ops port, dashboards, alerts, and where to look when something is wrong — read Observability.

5. Know the governance controls

As a platform admin you also operate the tenant-lifecycle and data-governance machinery:

  • Suspend an organisation to freeze its writes while keeping reads and data-subject exports reachable.
  • Offboard a suspended organisation — a previewed, confirmation-gated flow that exports then tombstones the tenant.
  • Data-subject requests (access and erasure) run against a personal-data registry that CI keeps complete.

These are summarised in Security & governance.

Next steps