OCR
Multi-channel intake: five channels, one single flow
Automatic invoice intake removes the step where somebody gathers the documents. Dedicated mailbox, upload, SFTP, API, scanned post, supplier portal, EDI (electronic data interchange) and e-invoicing platform all feed the same flow; OCR (optical character recognition) then comes in once, whatever route the document took, and nothing waits in a personal mailbox.
The problem
The flow starts before the tool
An invoice can arrive from five places at once, and it does in nearly every organisation: an accountant mailbox, paper post, the portal where the supplier uploads it themselves, a structured feed and the e-invoicing platform. Before reading a document, you have to have it. For the rest of the journey, see the full chain, from OCR to evidence.
The weak point is therefore not the reading, it is the collection. An invoice received in a personal mailbox waits for a return from leave. A document uploaded to a portal is only seen if somebody logs in. Processing time is decided there, not in extraction.
The cost of that dispersion is rarely measured, because it appears nowhere. It is spread across chasers from suppliers who were not paid on time, discounts lost for not having seen the invoice early enough, double payments when the same document arrived by two routes, and hours spent looking for a document nobody knows the entry point of. None of these items sits on a budget line, and none of them is the fault of one team in particular.
And what does not arrive in the same place is not controlled to the same level: the scope genuinely verified shrinks to whatever comes through the main channel. The other channels produce invoices handled from memory, with no trace of the date they arrived.
The arrival channels, and what unifies them
Every channel has its own constraints. They all lead into the same queue, with the same deduplication rule.
- Dedicated mailbox
- One address per entity or per site. Message body, attachments and download links are handled; the sender serves as a supplier clue.
- Upload and SFTP
- Manual drag-and-drop upload, or automated drop into a watched directory for bulk sending and overnight processing.
- Scanned paper post
- Post digitised on the copier or the production scanner comes in like any other channel; batches are split on arrival.
- Supplier portal
- Documents uploaded by the supplier themselves are retrieved without re-entry and attached to the party already known in your reference data.
- EDI and e-invoicing platforms
- Structured feeds and electronic invoices arrive in their original format: there is nothing to extract, only something to check.
- API
- The API (application programming interface) lets an existing system post a document and retrieve the extracted data, with no human intervention.
- Deduplication at intake
- The same invoice received by email and then by post is handled once: matching works on the number, the party, the date and the amount.
- Time stamping per channel
- Every incoming document is time-stamped with its arrival channel. The receipt date stops being an approximation rebuilt afterwards.
One channel deserves distinguishing from the others: the e-invoicing platform. The document arrives there already structured, so there is nothing to extract. What remains to be done is exactly what structuring does not settle: checking the values received against your negotiated terms, the purchase order and the goods receipt. Electronic invoicing removes the keying; it says nothing about whether the price is right. The same holds for an EDI feed, which carries a supplier price without carrying your contract.
One single flow means one queue of exceptions, one count and one audit trail, whatever route the document took.
What it changes
The delay shortens where it is actually created.
- Nothing sleeps in a personal mailbox: intake no longer depends on one person being available.
- The controlled scope becomes every invoice received, not the ones that happened to use the right channel.
- The receipt date is time-stamped per channel: payment terms are calculated on data, not on an estimate.
- Cross-channel duplicates are set aside at intake, before entering the accounts twice.
Unified intake is also what makes the count legible: one flow, therefore one counter of pages processed, whatever the arrival channel. Your ERP (enterprise resource planning system) remains the source of truth, and the documents received are returned to it with their receipt date and their channel, with no development added. Nothing has to be asked of the suppliers, and no existing channel has to be switched off before you start.
Capture
Gather and consolidate all your procurement data.
Collection by dedicated mailbox, upload, SFTP, scan, API or ERP export. Reads every format (native PDF, scan, photo, spreadsheet, structured feed), then extracts line by line: supplier, references, quantities, unit prices, discounts, taxes, terms.
The end of re-entry. A procurement history you can finally query.
What Capture detects
- Long documents and multi-page tables
- Unknown layouts
- Several languages and currencies
- Poor scans and handwriting
The subscription covers extraction and every intake channel. Checking your negotiated terms is the next step and is sized with you: what processing your volumes costs is shown tier by tier.
Nothing is lost. Everything can be checked, everything can be proven.
- The contract clause and the invoice line, highlighted side by side.
- Every extracted value stays linked to the exact place in the document where it was read.
- The same case produces the same decision, today as in six months: the rules are applied deterministically.
- No discrepancy is set aside in silence. Anything that matches no rule is raised, with its reason.
- The agent records what it did, in the order it did it: who, what, how much, when.
Frequently asked questions
Do suppliers have to change the way they send documents?
No, that is the very point of multi-channel intake: the supplier keeps sending exactly as they do today, and the document joins the same flow whatever route it takes.
How is the same invoice received twice handled?
It is matched at intake on the number, the party, the date and the amount, then marked as a duplicate. The first copy received stays the reference; the second remains viewable.
Does an electronic invoice go through the same extraction?
No, and there is no need: a structured feed does not have to be read. It joins the same circuit at the next step, the check, with the same deduplication and the same audit trail as documents received as PDF or on paper.
Which tools does Zylio connect to?
To the ERPs and management tools already in place, including SAP, Sage, Oracle and Pennylane, as well as to existing document repositories, mailboxes and feeds.
Measurable impact in every environment
More than 5 million procurement documents analysed
Between 1 and 7% of margin recovered
on the scope analysed
From 15 to 45% of time given back to teams, per FTE
depending on the scope and on data maturity
Zylio fits into your existing ecosystem.
The ERP runs the process. Zylio handles the exception and recovers the value that escapes it: invoices without a purchase order, line-by-line price discrepancies, duplicates and overbilling, off-contract spend.
Your data under high security.
Zylio meets the most demanding standards, and nothing is committed without your approval.
- Certifications
- Hosting
- Encryption
- Access
Continue on OCR
- Document extraction APIIntegrating without a project: API, exports, webhook, dedicated mailbox.
- Splitting batches on arrivalTwenty invoices in one PDF, twenty separate documents.
- Handling a supplier invoice, field by fieldThe fields expected, and the check that comes next.
- Multilingual and multi-currencyDate, thousands separator, currency symbol and tax mention: everything is normalised on output, the original value kept beside it.
See what this looks like on your own data
Twenty minutes, on a spend category of your choosing. We show you what the agents detect, with the evidence behind it.
- No commitment, on your own data
- Result in 3 weeks
- 20 minutes, no sales pitch
- Your data stays hosted in France
- No change of tool or process