01
Multi-tenant from row zero, not bolted on
Every tenant is fully isolated — schema, secrets via Supabase Vault, decision_log, kill switch. Switch tenants in one click without ever leaving the same Magistry surface.
// for agencies
Scale brands without scaling headcount.
Multi-tenant from the schema up. SSO, RBAC, per-tenant policy, per-store config, and an agency-of-record audit trail your clients can subscribe to. One operator runs ten brands without making ten exceptions.
Multi-tenant by design SSO + RBAC Per-tenant kill switch
// why agencies pick magistry
The operator-to-brand ratio your AOR pitch has been promising.
Headcount-light, audit-heavy. Magistry lets your strategist hold ten tenants the way they used to hold three — without changing what they get to see, sign, or revert.
02
SSO, RBAC, and per-tenant policy out of the box
Strategist sees everything. Junior buyer sees only the tenants and actions you grant. Client sees the digests you publish. One identity layer; permissions you can defend.
03
Per-store config — same agents, different rails
Apparel tenant runs Catalog tight, refund ceiling low. Beauty tenant runs CS open, frequency caps wide. Same Magistry. Different policy files. Zero forks.
04
Agency-of-record audit your client can subscribe to
Every external write is a row in decision_log with your operator stamped on it. Publish the read-only feed to your client and the 'what did the agency do this month?' deck writes itself.
// this week
What Magistry did across your tenant book this week.
A real week from a Magistry-powered agency running 11 brands from a single operator console — every row scoped to a tenant, every escalation routed to the right human.
- Ran the Catalog + Campaign + CS loop across 11 tenants from one operator console — 31,402 decision rows, 4 escalations queued to a strategist, zero cross-tenant leakage.
- Onboarded a new beauty client in 38 minutes — SSO connected, RBAC scoped to 'paid only', dry-run for 7 days, then flipped Campaign Specialist live on a 25% budget ramp.
- Published 11 weekly client digests directly from decision_log — every row linked, every change reversible, zero hours of agency time spent on the deck.
// from your dashboard
One operator. Eleven tenants. Eleven sets of rails.
Below: the agency operator console — every tenant a row, every row its own kill switch, policy file, RBAC scope, and decision_log link.
operator: jess.r (strategist) ── tenants 11 ───────────────────────────────────────────── tenant ks cs decisions(7d) esc linen-house OPEN OPEN 4,201 0 reef-and-range OPEN OPEN 3,108 1 northwood OPEN OPEN 2,884 0 solace OPEN HOLD 1,902 2 ← stoke-goods OPEN OPEN 5,210 0 ember-and-co OPEN OPEN 2,442 0 modular OPEN OPEN 3,710 0 brevard OPEN OPEN 1,840 1 plainfield DRY DRY 612 0 (onboarding) okay-wild OPEN OPEN 2,994 0 yarrow-and-iron OPEN OPEN 2,499 0 ───────────────────────────────────────────── totals 31,402 4 ks = kill switch cs = cs specialist DRY = dry-run only HOLD = held by you
// what you actually do
Pick the tenant. Read the four escalations. Move on.
You're not reviewing 31,402 rows. You're reviewing the four flagged for your call — and publishing the 11 weekly client digests from decision_log in one tap each.
otto: 4 escalations across 3 tenants · 11 client digests ready to send →
// what agencies are saying
From an AOR that finally stopped hiring per logo.
“You don't scale a brand by stacking headcount. You scale it by giving the rails to a teammate that never sleeps and always logs. We added four clients this quarter without adding a single strategist — and the client digest writes itself.”
Jenna Park
Director of Growth, agency-of-record
NORTHWOOD
// scale without staffing
Add the next five logos without hiring the next five strategists.
Spin up your agency tenant, invite your team with SSO, onboard your first client in under an hour, and publish the audit-trailed client digest the same week.
Kill switch ON by default · Dry-run from day one