Worked example
Invoice processing automation
You lose a day a month matching invoices to the bank statement, line by line. Most of it is boring and correct. The interesting line, the one that does not match, is buried in the twenty-three that do.
What changes
The workflow matches what it can and flags only what it cannot. Instead of reading every line to find the one that is wrong, you open a short exception list with the reason each item was held. A day a month becomes about an hour, and the hour is spent on judgement rather than ticking.
22 of 23
lines matched, VAT-checked and posted without anyone ticking them off. The twenty-third is the one that needs you.
How it works
01
Reads invoices and the statement
Line items, dates, amounts and VAT, pulled from PDF receipts as well as tidy exports.
02
Matches, checks VAT, posts
Clean matches post automatically with the correct nominal code and VAT treatment.
03
Flags only the exceptions
A short list of the lines that need eyes, each with the reason it was held.
What a month looks like afterwards
May reconciliation · 22 / 23 matched
1 exception to review
- Matched: 22 lines · €41,830.20
- Exception: Northline PO-00382 · invoice €1,240, statement €1,340 · €100 delta
- VAT totals: 23% band OK · 13.5% band OK
Illustrative output, not a client record.
What it plugs into
- Xero / QuickBooks / Sage
- Bank feed / CSV
- Dext / Hubdoc
- Google Drive
Approvals stay with a person
Nothing here removes the accountant. The tool does the reading and matching; a person still approves what matters, and every action is logged so an auditor can follow it. That is the same human-in-the-loop pattern used across every Artellis build, and it is what the EU AI Act expects. More on approvals and audit trails.
This page covers one workflow. How AI consulting runs here explains the wider six weeks: how a workflow gets picked, scoped, built and handed over.
Common questions
Does it work with PDF invoices, or only spreadsheets?
PDFs and photographed receipts as well as CSVs and spreadsheets. Most of the pain lives in the messy formats, so those are the ones it has to handle.
Will it post entries without me seeing them?
Clean matches post automatically with the right nominal code and VAT treatment. Anything ambiguous is held and flagged with the reason. You review the exceptions, not the whole pile.
Which accounting systems does it connect to?
Xero, QuickBooks and Sage are the usual ones, alongside a bank feed or CSV, Dext or Hubdoc for receipt capture, and Google Drive or SharePoint for filing.
How long does it take to build?
A working prototype on your own data in week one, and a live workflow inside about six weeks. Reconciliation is one of the faster builds because the right answer is easy to check.
What about VAT?
VAT bands are checked as part of the match and totals are reported per band, so a wrong treatment surfaces before it reaches your return rather than after.
Bring last month's statement
Twenty minutes is enough to tell whether your reconciliation is a good first workflow or whether something else in the business is costing you more.