Purchase orders, delivery notes and invoices compared in seconds. The system flags every price or quantity discrepancy — with or without AI.
Request free demo
Manually verifying the consistency between orders, delivery notes and invoices is repetitive, slow and error-prone. A wrong price or a mismatched quantity can mean incorrect payments, disputes with suppliers and hours of wasted work.
The unit price on the invoice does not match the agreed order.
The invoiced units do not match those delivered or ordered.
A line present in the order does not appear in the delivery note or invoice.
The VAT rate applied does not match what was expected.
Choose the comparison type, select the AI provider and upload the documents: order, delivery note and/or invoice, plus the payments if you also want to check the settlement.
The system extracts lines from the documents, compares them and classifies anomalies in the background.
View the report with detected anomalies, their severity and a direct comparison of values between documents.
Compare the received invoice with the delivery note. Detect quantity and price differences between the two.
Verify that the invoice respects the prices and quantities of the original order sent to the supplier.
Upload all three documents. The system compares the entire document chain in a single session.
Each completed session generates a report with the comparison type chosen, the AI provider used and the reference supplier. The summary immediately shows how many errors and warnings were detected.
Units in one document do not match those in another. The system shows the values side by side for immediate comparison.
The unit price on the invoice does not match that of the order. The percentage gap is indicated.
The VAT rate applied on the invoice does not match that expected from the order or delivery note.
A line is present in one document but not in the others. This may indicate an item invoiced that was not in the order.
An expected line based on one document does not appear in another. Possible omission on the invoice or delivery note.
Comparing order, delivery note and invoice tells you whether the invoice is correct. It does not tell you whether it was settled, for the right amount and on time. Tick one box and you can upload the payments too — a bank statement exported from your online banking, a list of movements or a transfer batch, as PDF, CSV or Excel — and the system matches movements to the invoice by reading the description, the amount and the supplier. Then it adds up what was paid and checks that the figures balance.
No movement can be matched to the invoice. The simplest finding, and the most expensive one when it slips through.
The matched movements do not cover the total. The report states how much was paid and how much is still missing.
More went out than was due. This also happens when a single transfer settles several invoices: the difference is shown, not hidden.
The balance arrived after the due date on the invoice, beyond the tolerance days you decide.
The other operations in the bank statement are reported in a single summary line, so the report stays readable.
Invoice total, amount paid, difference, due date and settlement date in a single panel. And it exports to Excel with numbers ready to compute.
A PDF that comes out of a scanner is an image: it holds no text to read, and automated checks stop there. The system notices on its own and, if you authorised visual reading, hands that document to an AI model able to read it the way you would. You declare nothing at upload time.
The system works out which documents are scans. You upload the files as always.
PDFs with text are processed as before, at no extra cost. Visual reading kicks in on scans only.
The feature is switched on with a tick and has a page cap per document: nothing is consumed without you knowing.
Each completed session shows a summary with the number of lines compared and anomalies found, classified by severity. For each anomaly, the type, percentage difference and lines involved in both documents are shown.
The module integrates with the main LLM providers. Each tenant configures and manages their own provider directly on the platform — AI costs always remain under your control.
Book 30 minutes: I will show you AI Document Checker on your own documents and we will see together whether it fits.
Request the free demo
Terra di App
Custom cloud business software, built on a base that already works. Source code and infrastructure stay with the client.
2026 Dino Solano · VAT 10312320962 · Milano · All rights reserved