Built by a retailer, for retailers.
Built by retailers, for retailers. Thirty years from every side of the table: the vendor shipping in, the distributor in the middle, and the retail head office deciding what goes on the shelf. The first version of this engine ran across hundreds of stores for a major retailer. What runs today is a ground-up rebuild: the same understanding of the problem, on hardware and models that did not exist when it was first written.
Building it is six weeks. Knowing what to build is thirty years.
Anyone can tell you they understand retail. Here is what understanding it looks like: the things a system has to know before the maths matters.
Your system orders in packs of 12. Your supplier changed to 6 in March.
Nobody told the system, so every order since has arrived half short. And separately the on-hand says 14 when there are 3. Shrink, a double-scanned receipt, a case on a pallet in the back. A model that assumes clean data doesn't fail loudly; it produces confident nonsense, which is worse.
A line out of stock for six weeks looks exactly like a slow seller.
Most systems read the flat line and cut the order, so it stocks out again, so they cut again, quietly strangling the line they should be growing. I know because it happened to me. Twice.
Retail runs on a calendar, not a model.
The order goes to the supplier Tuesday. The DC picks Thursday. Review cycles, lead times, per-store run hours, month-end. Get the rhythm wrong and the answer arrives after the decision.
Sometimes the right answer is "not enough data to say."
Every competitor's demo answers everything. This one is built to say "I don't have enough here to give you a credible number", and to tell you what to fix so that next week it can.
Vendor. Distributor. Retailer. Then a developer with a grudge.
Vendor side, shipping in. Years of trying to get a straight answer out of a retailer's data. Sell-in versus sell-through, and a relationship built on whichever of the two the other side preferred to quote.
Distribution. The middle of the chain, where inventory gets stuck and where the question "whose fault is the stockout?" is unanswerable without the numbers nobody had.
Retail: head office, store operations, finance. Inside a national chain and a global one, watching good teams make decisions on incomplete data because the tools that could help cost six figures and took a year.
The engine. The first version was built for a major retailer, across hundreds of stores, in production, fifteen years ago. It worked, and then it waited, because doing this properly meant running the model inside the retailer’s own building, and that was not possible at a sensible price until recently. What runs today is a ground-up rebuild, and it thinks for itself: to forecast, to order, to explain every recommendation in a sentence, and to ask permission before it does anything on its own.
"Will you still be here in two years?"
It runs in your building, which means it is not a service that can be switched off from somewhere else.
A cloud vendor going quiet takes your system with it. A box in your building does not stop working because a supplier had a bad year. That is the honest answer to founder risk, and it is a structural one, not a promise.
The other honest answer: I don't have a reference site yet. You'd be the first, and I'd rather say that now than have you find out in week three. That is a real thing you would be absorbing, and it is priced and written down as such, in what you pay and in what I commit to in writing. Founding terms exist because I don't have a reference site yet. They end when I do.
They probably can. The code is the easy part now.
Someone will offer to build you this with AI. What they cannot do is know that a line out of stock for six weeks looks exactly like a slow seller, and that every system that misses it quietly strangles the line it should be growing.
The question is whether you want your team spending the next two years discovering the exceptions I have already been beaten up by, or spending it on something only they can do. We are happy to talk about which part you are actually buying.
Rob · Founder, RetailPulse · Johannesburg · hello@retailpulse.co.za