AI built around how a wood products plant runs.
Rough and kiln-dried lumber into finished product: flooring, mouldings, window frames, joinery, pallets, panels and engineered wood. Where the order book meets real capacity.
Real numbers from real operations.
Every figure traces to a working paper. Where a number is unproven, we say so.
Four phases, from the order taken to the margin earned.
This is the most varied link in the chain, from a two-moulder shop to a panel line. The cycle below is the same in all of them, and so is the place the information breaks down.
Sales & Receipts
Quote · order · ship · receipt
Every rep runs their own book, and nobody sees all of them together until somebody rebuilds the month in a spreadsheet. A rep's best month on revenue can be their worst on margin, because discount, freight and the product mix they happened to sell all land somewhere else. By the time that shows up the quarter is priced.
Quarri gives every rep a live book with revenue and margin side by side, so the mix that actually earns is visible while the quarter can still be changed.
- Every rep's book live, not rebuilt each month
- Revenue and margin ranked side by side
- Receipts tracked against what was shipped
- Customer and product mix visible per rep
The biggest book is not always the best one. Volume and margin rarely rank the reps in the same order.
Procurement & Margin
Source · buy · receive · benchmark
Buying runs across a long tail of suppliers, priced in different units on different terms, so almost nothing is like for like. Without a comparison nobody can say whether a price was good, only whether it was normal. Price is only half of it: a supplier who is cheap and late costs more than one who is dearer and dependable, and almost nobody scores that. At one operation six in ten purchase orders were arriving late and nobody had ever measured it.
Quarri compares every purchase line against your own buying, scores each supplier on promised against actual, and matches the order to the receipt and the invoice in the same pass.
- Every purchase line compared with your own buying
- Suppliers scored on promised against actual dates
- Purchase, receipt and invoice matched per line
- Unit and conversion errors caught, not carried
- Input costs tracked against published price indices
A single unit-of-measure error had inflated purchase spend for years and turned a healthy margin into a reported loss.
Inventory
Stock · reman · lead time · cover
Inventory usually gets managed backwards: you look at what you hold and react to it. What actually decides whether you meet next month's orders is what that stock can become and how long each path takes. Which items can be remanufactured into what, how a reman run compares with a fresh production run, and what your suppliers' lead times really are rather than what their terms say. Without that, stock is either short or dead, and both cost you.
Quarri holds one live position across stock on hand, reman routes, production and reman times and real supplier lead times, so cover is calculated against the orders you actually have to meet.
- On-hand stock joined to what it can become
- Reman routes and run times held per product
- Supplier lead times measured, not taken from terms
- Cover calculated against every open purchase order
- Ageing stock flagged well before it goes dead
Cover for the open order book, built from what you hold, what it can be remanufactured into and what is inbound. The gap shows while there is still time to close it.
Scheduling & PO Optimisation
Schedule · route · promise · deliver
A purchase order is a promise with a date on it, and the plant has several ways to meet each one. The difficulty is that they compete. The rough stock that would fill one order fastest is the stock another order needs, and the moulder that could run both is only free on Thursday. Scheduled one at a time you hit most of your orders. Scheduled together you can hit all of them.
Quarri takes the live inventory position and routes every open order through the plant together, so the schedule you run is the one that hits every purchase order rather than the one that hits the loudest.
- Every open PO scheduled together, not one at a time
- The live inventory position drives what can be promised
- Routes chosen so orders stop competing for stock
- Same-product runs sequenced into one setup
Four open orders sent down four different paths so none of them competes for the same stock. All four land inside their promise date.
AI at the foundation. Not retrofitted.
Most software for this chain was built as a form. Quarri was built as a model of your operation.
- Their fields, your words mapped in
- Exceptions become change requests
- Paper is a keying job
- Every run costs a model call
- New questions join the queue
- Reports you wait to receive
- Scale by hiring around the gaps
- Your words, from day one
- Exceptions become rules you edit
- Paper read at volume, deterministically
- 100x cheaper per execution
- Questions asked in plain English
- Dashboards and scripts you build yourself
- Scale by pointing it wider
Ask it in plain English. Across all four phases.
Because the sales book, the purchase ledger, the stock position and the schedule all land in one governed layer, a question can cross the whole plant instead of stopping at a system boundary. That is a margin lever, not just a tidier way to operate.
- Any question across the cycle, in plain English
- Numbers computed, so the answer matches the ledger
- Dashboards and scripts your team builds and shares
- Alerts when a SKU stalls or a margin slips
Margin per linear foot after landed cost and setup. The question crosses four systems; the answer does not care.
Wood products is not one business. We know which one is yours.
The largest and most varied link in the chain, from a two-moulder millwork shop to a panel line. The platform is the same; the vocabulary, units and workflows are yours.
Mouldings, flooring and trim
S2S and S4S programmes, T&G and groove stock, random-length runs priced properly, and grade recovery from FAS to Economy tracked per moulder.
Pallets, skids and crating
Heat-treated stock certified and tracked, cut-stock yields per work order, and repair-versus-new economics visible by account.
Stile & rail, frames and stairs
Judgement-heavy works orders modelled, component yields tracked input SKU to output SKU, and the one critical workbook automated.
OSB, MDF, glulam, CLT, LVL, I-joists
Furnish and resin cost against panel margin, press utilisation named honestly, and dispatch sequenced against real curing time.
A lumber and millwork manufacturer
Purchasing ran across a long tail of suppliers with no like-for-like comparison, stock was reported as a total rather than a position, and every sales rep's book was rebuilt by hand at month end.
Quarri matched purchase, receipt and invoice line by line, surfacing a six-figure procurement saving and tracing a unit-of-measure error that had inflated purchase spend for years. Reading the full remanufacturing ledger showed set-up, series and knife changes were 46% of mill cost against 28% for the ripping that actually adds value, which is a sequencing problem rather than a machining one.
Agents that order, re-promise and chase.
Everything above works out the answer and hands it to you. The next step is letting it act, on the thing that cuts across all four phases: cash. An agent that raises the purchase order when cover runs short against the open book, re-promises the customer when a run overruns, and chases the invoice that has quietly gone past terms. You approve a decision instead of assembling one.
- Purchase orders raised when cover runs short
- Customers re-promised when a run overruns
- Receivables chased the day they pass terms
- Every action proposed for approval, never taken quietly
A run overruns and the promises around it are re-cut. A design sketch rather than a screenshot: not deployed today, and every action would come to you for approval.
Most operators sit on more than one link.
See it on your own data.
Live in two weeks, on the systems you already run.