What Fleet solves

A large enterprise has dozens of business units, each with its own compliance posture. Finance is different from Engineering. Canada is different from the US. A shared services group operates AI across them all. "One policy fits all" does not work; "everyone defines their own" does not audit.

Fleet introduces a tenant hierarchy. Global policy rolls down; unit-level policy overrides within the bounds the parent set; the unit cannot relax a rule its parent enforced. Every override emits a signed event. Every budget is per-tenant.

Tenant isolation and policy inheritance

  • strict isolation: one tenant's evidence never visible to another; per-tenant Ed25519 signing keys
  • inherited policy: parent rules carry unless the child has an explicit allow or reroute
  • ratchet-only relaxation: a child can be stricter than a parent, never looser
  • per-tenant spend caps: budget enforcement composable across the tree
  • cross-tenant reporting: the Fleet console aggregates without exposing raw evidence

Deployment postures, same guarantees

Fleet supports every RelayOne posture in a single tree. One business unit runs on-prem; another runs sovereign BC-Canadian; a third runs hybrid. The parent sees policy compliance uniformly, because the integrity story does not depend on the posture.

Shift-handoff dashboards

Purpose-built for SOC teams doing shift changes. One view, one timezone, exportable. Includes: open policy exceptions awaiting review, reroute-rate spikes since last shift, budget threshold breaches, tenants that flipped posture, flagged identities.

Attach the CSV to the shift ticket; the next shift picks up with the same context.

RelayGate edge, per-request RelayOne fleet control tenant: engineering hybrid tenant: finance on-prem tenant: ca-health sovereign BC evidence bundles
RelayGate sits in front. RelayOne governs the fleet. Tenants run in whichever posture compliance requires. Evidence bundles aggregate at the fleet level.

Adjacent reading