SG Schema. The operational blueprint behind every SimpleGrid deployment.

SG Schema is the artefact AI writes from a 3-hour conversation with your founder and ops head. The engine reads the schema and generates a working custom ERP. No predefined modules. No template configuration. This is how SimpleGrid ships a custom system in 7-21 days at our risk.

What it is

SG Schema is a structured, machine-readable description of how a specific factory actually runs. Not how the brochure says it should run. The actual stages, the actual contractors, the actual approval chains, the actual costing logic, the exceptions every floor has.

An SG Schema names:

• Entity roots - the nouns your operation cares about (job, lot, batch, order, dispatch, vendor, customer, machine).
• Stages - the verbs that move an entity from intake to dispatch (received, cut, stitched, packed, QC'd, shipped).
• Approvals - who has to say yes before what moves where, including the exception path when they don't.
• Costing logic - how material, labor, machine, and overhead get attributed per entity, per stage.
• Roles + permissions - who sees what, who can edit what, who can override what.

Two factories making the same product can ship completely different schemas - because they run differently. That's the point.

How it gets written

A 3-hour live call with your founder, ops head, and a couple of floor leads. We map your workflows live, asking the questions only an operator would think to ask. AI drafts the schema in real time during the call. By the end you have an operational blueprint of your factory - in your language, not vendor jargon.

This is the same blueprint that drove the 21-day Furniture Exporter deployment and the 12-day Apex Apparel deployment. Two completely different schemas, one engine, both built around the operation - not the other way around.

What the engine does with it

The SimpleGrid engine reads the SG Schema and generates a working custom ERP:

• Database tables, indexes, and event-ledger streams shaped to your entities and stages.
• Forms, lists, and dashboards generated per role - operator, supervisor, plant manager, founder.
• Approval flows wired to the people you named, not generic placeholders.
• Costing rollups computed exactly the way you described them on the call.
• An AI floor-staff chatbot that maps plain-English actions ("received 200 sheets of steel from Midwest Supply") to the right structured event in your schema.

Why it changes the timeline

Module-based ERPs (NetSuite, SAP, Epicor) ship a predefined data model and ask you to bend your operation to fit. That's why a typical mid-market ERP takes 12-18 months and three failed attempts before something runs.

Schema-driven means the data model is generated after we know how your factory runs. The 90% of deployment time that usually goes into "discovery + configuration + customization" gets compressed into a 3-hour conversation. The remaining work is the engine generating, our team reviewing, and you running it.

The contrast

• Module-based ERP: you adapt to the system. 12-18 months. $500K+ minimum. Three implementations in to find the one that fits.
• SG Schema-driven ERP: the system adapts to you. 7-21 days. We carry the cost. You pay if it moves the business.

Read deeper