Two jobs: classify every spare part A / B / C, then set the right MIN and MAX for each one. Parts are ranked on unit value × quantity checked out, split by cumulative Pareto share, benchmarked against what a healthy store normally looks like, then given a reorder point (MIN), an order-up-to level (MAX) and a stocking strategy — computed in the page, reviewable by the AI model of your choice.
Columns are matched automatically by header name (English or French). Check them before reading the result — anything can be re-pointed. The template holds the exact headers this app reads, with three filled example lines; the workbook this app exports can be loaded straight back in.
Export the levels, review or overrule them in Excel with the team, load the file back: the MIN and MAX columns you changed replace the computed ones everywhere — table, KPIs, workbook and the push to the CMMS. The template carries the same columns with one filled example.
The numbers above are computed inside this page and never leave it. A review sends only the parts you select — code, class, demand, batch, lead time, value and the computed baseline — and asks the model to confirm or overrule the min/max and name the stocking strategy. No API key at the client? Copy the prompt into any chat and paste the answer back.