Manufacturing ERP, built to your floor.

On site for the parts that need it, remote for the rest. The discovery week happens in the plant. We walk the line from goods inward through to dispatch, stand at the machines while a shift changes over, and read the registers people actually fill in rather than the ones described in the office. A job card cannot be designed from a video call, and the distance between how a plant is described upstairs and how it runs at the press is usually the whole project. Go-live is in the plant too, because that is when a supervisor asks the question nobody thought to ask in a meeting. The build between those two points is remote, on a weekly call with something working to look at each time. If you need someone in the building every day, we are the wrong team, and we will say so on the first call rather than three months in.

Sometimes you should take the package, and you will hear that from us after we have seen the floor rather than before. Packages are strong on what every plant shares — purchase, stores, invoicing, statutory returns — and weak on the part that is specifically yours. Watch how a demo handles job work: material issued to a subcontractor, returned part-consumed and part-rejected, costed against a job still open. Ask what it does with a lot split across 2 machines, or a customer variant that changes the routing without changing the item code. If either answer involves a spreadsheet kept alongside, you have found the gap, and the gap is where a register reappears on the floor and you end up with 2 versions of one number. Our test is whether the process that earns your margin fits the package as sold. If configuration cannot close that gap, no amount of it will.

Within weeks, on one process at a time, while everything else carries on as it does today. We take a single process that can be checked end to end without depending on the rest — usually attendance, stock issue or job cards — and put it into real use at the machines and counters it belongs to. For the first month it runs beside whatever you do now, so the two can be compared at close and nobody is asked to trust a screen over a register they have kept for 15 years. Supervisors learn one screen rather than a system, which is the difference between a rollout people adopt and one they quietly work around. When the numbers agree for a full month, the register stops and the next process starts. Anyone offering a single date on which the whole plant changes over is describing the scenario behind every ERP horror story you have been told.

Yes, and that is usually where the quickest return sits. ESSL biometric readers are what we run in production today: the punch log feeds shift, overtime and leave rules and closes the month into a file payroll accepts, with no export-and-repair step in Excel in between. A weighbridge ticket can post against the purchase order it belongs to instead of being written on a pad and keyed in an hour later. Scanning at issue and at dispatch removes the second entry that creates the difference between physical stock and system stock. Machines exposing a counter or a log can be read directly. Machines that do not still need a person to enter output, and we would rather build a 2-tap screen at the machine than pretend otherwise. Accounting normally stays where it is — Tally or similar — and receives what it needs instead of being replaced.

It keeps running, because nothing in it depends on us being reachable. The system sits on servers or a cloud account in your company's name, the database is yours, and the code is in a repository your organisation owns from the first week rather than the last. Readers on the wall keep reading, job cards keep printing, and the month closes the same way whether or not we answer the phone. We write the runbook for the dull emergencies — a disk filling up, a device dropping off the network, a restore from last night's backup — in language your IT person or your nearest competent vendor can follow at 2am without calling anybody. There is no lock-in and no retainer you cannot end. The only thing we would ask for is a week of overlap with whoever takes over, which costs us little and saves them a month of reading.

One week and one module. The week is discovery, on your floor, and it ends with a written scope and a fixed price for the first piece of work — a price, not a range. That first piece is chosen to be small enough that being wrong about it is survivable, and useful enough that you would want it even if the rest never happened: job cards at one section, or stock issue at the stores window. You commit to that and to nothing beyond it. If the price and the problem do not meet, you find out in week 1 rather than in month 6 with half a system installed and an annual licence already signed. We publish no rates, because a figure quoted before anyone has walked your floor is guesswork with a decimal point on it, and a single shift of 40 people is not the same job as 3 plants running rotating shifts.