Architecture & tech
Designing a Data Model for Multi-Plant, Multi-Entity Manufacturers
Companies, plants, warehouses, currencies and common masters — decisions that are expensive to change later.

KEY TAKEAWAYS
- Some data model decisions are very expensive to change later.
- Model legal entities, plants and warehouses separately.
- Keep masters shared and transactions scoped.
- Design permissions around the same hierarchy.
Software that works for one plant often struggles when a second company, a new plant or an overseas subsidiary appears. The cause is usually the data model.
Here are the decisions to get right early.
The hierarchy
| Level | Represents | Examples of scoped data |
|---|---|---|
| Group | Whole organisation | Common masters, group reports |
| Company (legal entity) | Books of accounts, tax registration | Ledgers, invoices, tax |
| Plant / site | Physical location | Production, machines, people |
| Warehouse / store | Stock location | Inventory, bins |
| Line / work centre | Production unit | Work orders, OEE |
Shared masters, scoped transactions
Items, customers and vendors are usually shared across the group, with company- or plant-specific attributes where needed (local prices, tax codes). Transactions always belong to a company and plant.
Currencies and units
Store transaction currency and company currency, with exchange rates by date. Store a base unit per item with conversions, never assume units.
Inter-company flows
- Stock transfers between plants of the same company.
- Sales between group companies with matching purchase records.
- Elimination for group reporting.
Permissions follow the hierarchy
Users are scoped by company, plant, warehouse or territory. Reports and AI queries respect the same scopes.
Frequently asked questions
Should each plant have its own item codes?
No — use shared item codes with plant-specific attributes.
How do we handle acquisitions?
Add a company, map its masters to group masters and migrate transactions in waves.
What about different tax systems?
Tax configuration belongs to the company level.
Can one user see several plants?
Yes, through roles scoped to multiple plants or companies.
- #manufacturing data model
- #multi entity ERP data model
- #multi plant database design
- #master data design
- #multi currency manufacturing
Trikaan Research & Engineering
Trikaan
Written by the researchers and engineers who build Anvaya, ASTR and Seed Track, with input from Trikaan's delivery and engineering teams.


