Apps / Gradora
Two tools for the same factory. One plans what to make and buy, and holds the plan to account afterwards. The other asks whether the process making it is stable, capable, and understood. Both run on a Mac at a desk or an iPad on the floor, both show their working, and neither one sends your data anywhere.
Nothing is a black box. Any figure on screen can be followed back to the row, the order or the measurement that produced it, and the rules that shaped it are written down rather than implied.
One engine, two destinations. Keyboard-fast on a Mac beside your spreadsheet, readable on an iPad beside the machine or in the meeting. iPhone is not a target for either.
No account, no server, no analytics, no sync. Production data and commercial forecasts are among the most sensitive things a business holds, and these apps have nowhere to send them.
They are separate apps, bought separately, and neither needs the other. They share a name because they share a way of working: real data in, a defensible answer out, and the reasoning visible the whole way through.
Sales, inventory and operations planning for a single site. A demand planner imports seven extracts from the ERP and CRM and leaves with a part-level action list: what to buy, what to build, how much, and by when, under three demand scenarios. The cycle is published as a record that cannot be edited afterwards, scored for forecast accuracy as the actuals arrive, and handed to the finance director as a meeting pack that needs no editing.
Statistical process control and designed experiments, for quality and process engineers. Control charts, capability studies, designed experiments and FMEA, from real measurement data, with every statistic coming from a tested function rather than a guess.
Because they are used by different people, on different days, to answer different questions.
Planning runs on a monthly rhythm and is owned by whoever has to stand up in the S&OP meeting and defend a number. Process control runs continuously and is owned by whoever has to prove a part is in tolerance. Bolting the two together would produce one large application that is worse at both jobs and forces every buyer to pay for the half they do not need.
What they share is the principle underneath: a number you cannot trace is a number you cannot defend, and software that hides its working is not doing you a favour.