Load by load. For each load, it reads three records: the load record made when the truck left the tract, the scale ticket from the mill, and the line on the buyer's settlement or stumpage statement. It matches them on the load number, or by date, truck and weight where the number is missing. It applies the contract's rate and conversion for that product, and compares the result with what was paid. Whatever doesn't match is listed with a likely reason, for a person to settle. Contractor invoices for logging and hauling are matched to the same loads.
What the usual answer says
Descriptions of AI reconciliation say it extracts data from documents, matches records across systems, flags discrepancies and replaces spreadsheets. That is the shape of it. It leaves out the level at which the matching happens, which decides what it finds.
The records already exist
A pay-as-cut sale is designed to create the records needed for reconciliation. A sample timber sale contract published by a university extension service in 2016 sets out the pattern. Before each load leaves, the buyer records "the load number, type of forest product, destination, date and time of hauling" in a logbook. "The load number will also be printed on each scale ticket and a copy of the scale tickets will accompany each week's payment", along with "a statement showing the volume of wood by weight and type removed from the property".
A 2024 extension guide by Brady Self, on timber sale agreements, suggests a scaling clause "could be useful" for pay-as-cut sales: "This may include designating log rule, volume table, or weight conversion factors to be used, and where, when, and how timber is scaled or weighed".
So a contract on this model puts the three records, and often the rules to connect them, in place. The work left is connecting them for every load, every week.
Why totals are not enough
A weekly settlement is a total. It can agree with the week's tickets to the ton and still be wrong. A load of sawtimber paid at the pulpwood rate, a ticket missing and another counted twice, a load delivered to a different mill than the logbook says: each moves money, and some cancel in total. Only a load-level match shows them.
By hand, a load-level check is slow, so it tends to be done on a sample, or when something looks off. Accounts payable teams in general know the cost of exceptions. Ardent Partners' 2025 AP metrics report, sponsored by the e-invoicing company Pagero and drawn from 212 respondents, put the average cost to process a single invoice at $9.40. In the same survey, 53% named invoice exceptions as a challenge, the most-cited of any.
Where the buyer generates the settlement from electronic scale tickets, load by load, the totals-versus-loads problem shrinks. The risk then moves to the rate table and the product coding: a product mapped to the wrong rate, a rate not updated when the contract changed. Those are what to check in that case, and they show up in the same load-level comparison.
What AI does in the process
It reads the three records, whether they arrive as scans, PDFs, spreadsheets or system exports. It matches loads, using the load number first and date, truck, weight and destination where the number is missing or wrong. It applies the contract's rate by product and any conversion from weight to volume the contract names. It lists every residual with its likely cause: rate difference, missing ticket, duplicate, weight outside the usual range for that truck.
From Quarri's own work with a forestry operation: one reconciliation cycle surfaced an over-accrual credit of $80k+.
Contractor invoices on the same loads
Logging and hauling invoices are the fourth record. Matching their lines to the loads, rather than to a monthly total, finds loads billed twice, loads billed at the wrong haul distance and loads that were never delivered. The same load number, or the same date, truck and weight, ties the invoice line to the ticket.
Deciding what each residual means, and raising it with the buyer or contractor, stays with the forester or the accountant.
When it doesn't apply
Lump-sum sales have no per-load payment to reconcile, though load records still guard against removal beyond the sale boundary. And the 2016 sample contract is one extension service's model; contracts vary, and the method follows whatever records the contract requires.
Quarri for forest management is built around how a forest operation runs, from the cruise to the settled account.
Sources
- University extension service, "Sample Timber Sale Contract", published 15 June 2016: content.ces.ncsu.edu
- Self, "Marketing Your Timber: The Timber Sales Agreement", university extension publication 1855 (POD-05-24): extension.msstate.edu
- Ardent Partners, "Accounts Payable Metrics That Matter in 2025": datocms-assets.com
- Quarri evidence ledger, E14 (proven)
Quarri is an AI-native data platform for the timber supply chain. It connects buying, production, sales and inventory for forest management, sawmill, wood products and pulp, paper and packaging operators.