Software that smells like a factory floor.
ControlGrid OS wasn’t designed in a product workshop. It was built inside a working food factory, screen by screen, because the factory needed it — and then it turned out every factory does.
The first user of ControlGrid OS was a manufacturer and distributor in Antwerp with a very normal set of problems: production planned in spreadsheets, orders arriving on WhatsApp, shop counts on paper, invoices sent when someone found an hour, and a bank account that was always the last to know.
Instead of stitching together six tools that each solved a tenth of the problem, the owner built the system he actually needed — one place where the whole week is visible: what to make, what to send, what to bill, what came back, who clocked in, and whether the money landed. Every screen was tested the same day it shipped, by the people doing the work it describes.
That factory still runs on ControlGrid OS today — from customer website and team chat to production, distribution, payroll and the bank. It is our flagship deployment, our proving ground, and the reason the product reads in plain language instead of ERP jargon: the first user demanded it.
Once the whole operation lived in one system, AI could become useful instead of generic. The standard ControlGrid guide is Mr. Grid; each company can make him its own, and Tasty Foods named him Mister Tasty. The name made him familiar; the connected data made him valuable. The team can ask a normal question, get an answer from the operation and teach him the unusual names and rules only their business knows. He can also prepare a draft sales order, create a client or update a shop’s distribution setup, with the real records resolved before a person confirms. Focused AI also structures spoken counts and supplier invoices.
Operators first
If a screen needs a manual, the screen is wrong. Plain words, big numbers, obvious next steps.
The week is the unit
Manufacturing runs in loops, not lists. The system is built around your weekly cycle end to end.
Calm is a feature
The dashboard tells you what needs attention and stays quiet about everything that doesn't.
AI earns its place
The assistant answers from real operations, remembers what the team explicitly teaches it, and shows a confirmation before a supported change.