Your silo, explained

← Back to Help Center

What is a silo?

A silo is a complete, isolated deployment of CashSheet for your organization. It has its own Postgres database, its own application containers, and its own subdomain. No other customer touches your silo's data — the isolation is enforced at the infrastructure layer, not in application code.

Concretely: if your tenant slug is blythe, your silo is everything that lives at blythe.cashsheets.com, plus the dedicated database cutsheet_blythe running on our Postgres cluster.

What gets provisioned

When you sign up as a tenant, four things spin up automatically:

1. Postgres database named cutsheet_{your-slug}, with two roles:
   • cutsheet_{your-slug}_admin — owns the public schema (the CashSheet web app uses this)
   • cutsheet_{your-slug}_daemon — owns the master schema (the CutSheet LP solver uses this)
2. Application containers for your CashSheet web app and your CutSheet daemon — one set per facility you operate.
3. Subdomain at {your-slug}.cashsheets.com, with a wildcard Let's Encrypt TLS certificate.
4. SSO secrets — HMAC keys that bridge your silo with accounts.cashsheets.com for sign-in.

All of this is provisioned asynchronously after your organization is created on the platform. You'll see the status field on your tenant record move from pending → provisioning → ready. Typical time from “ready” status to first user being able to log in: a few minutes.

The two web addresses you'll use

Most customers find this confusing at first. CashSheet lives in two places:

Once you've signed in at accounts.cashsheets.com for the first time, you can bookmark your silo URL and go straight there. The first login goes through accounts (so SSO can issue your silo session), but subsequent visits don't.

Multi-facility tenants

If your organization operates multiple facilities — multiple plants, distribution centers, freezers — each facility runs its own container set in the same silo. The data lives in one shared database, but each facility has its own:

• facility_id (e.g. BLYTHE, WAGYU-CA, COSMOS) — used as the atp_instance_id in the CutSheet daemon and as a partition key on most operational tables.
• Independent LP solver— each plant's LP runs separately with its own yield matrix, capacity constraints, and order book.
• Shared chart of accounts and financial reporting — when you close the books, the entity-level rollup pulls from all facilities. Inter-facility transfers post proper journal entries automatically.

You add new facilities from the platform admin (accounts.cashsheets.com → tenant detail page → Facilities). Same silo, new container set — provisioned in the background once you save.

Provisioning timeline

A typical first-time tenant onboarding runs two to three weeks end-to-end:

1. Day 1 — Silo provisioned (automatic, ~5 minutes after the tenant record is created). You can log in and see the UI; most tables are empty.
2. Days 2–5 — Master data load: chart of accounts, customers, vendors, products. Most of this can be CSV-imported from your existing system.
3. Days 5–10 — If you're using CutSheet: yield matrix, capacity constraints, price file, sample order book. We sit alongside your team for this.
4. Days 10–15 — Parallel period: run your real shifts in both your old system and CashSheet, compare results, tune the matrix.
5. Days 15–21 — Cut-over: live on CashSheet, old system becomes the audit reference.

Larger or multi-facility deployments take longer — typically four to six weeks. We'll give you a per-engagement timeline before you sign anything.