Invoices, reminders and other attachments land in the same mailbox. Automatic email classification sorts what arrives, sends invoices straight into processing and leaves the rest untouched.
Schedule a meeting
Finance teams receive invoices by email, mixed with quotes, reminders, correspondence and spam. Sorting by hand takes time and leads to missed invoices. The question is not whether errors creep in, but when.
The eConnect Email Receiver analyses every incoming email as soon as it arrives. Classification combines a visual fingerprint of each document with textual checks via OCR. Emails that contain an invoice (PDF or XML) are forwarded into processing: PDF invoices go to the AI-driven Intelligent Document Recogniser (IDR) for Scan & Recognise, XML invoices go straight through the regular processing flow. Emails without an invoice stay in the regular inbox.
Email classification is more than recognising attachments. The system assesses the full email and routes each part to the right destination.
Per email address you set which senders are allowed. Four levels: only verified users in the organisation, everyone in a specific domain, one specific email address, or open receipt without sender validation. An optional organisation filter further limits which supplier may act as sender. Emails from senders that are not allowed are rejected with an error message.
<kvk>@inkoopfacturen.econnect.euThe Email Receiver fits the existing email infrastructure, without suppliers having to change address.
Organisations with several administrations in one environment do not have to give suppliers a separate email address per administration. Administration detection determines which administration an incoming invoice belongs to. Identifiers (VAT number, KVK number) take priority. Without an identifier match, the system looks at organisation name, city and postal code. An exact match assigns the invoice immediately. Several possible matches send the invoice to Quality Control. This requires at least a Professional subscription and is named separately in the contract.
Classification distinguishes document types including invoices, purchase orders, pro formas, receipts, packing slips and reminders. Each type gets the right route: invoices go into processing, packing slips and specifications are attached to the invoice, reminders and pro formas are not processed.
In the platform you follow processing status per document, from receipt to delivery in your accounting software. Documents the system cannot classify with certainty are reviewed by the Quality Control team; you then see the reason for the message. If an invoice is classified as the wrong document type, report it to support and the classification model is improved for later documents. Read more in the documentation on document classification.
Classified PDF invoices are turned into validated e-invoices by the AI-driven IDR.
Enrich recognised invoice data with cost centres and general-ledger accounts via business rules.
The full chain from receipt to a booking proposal, for every channel and format.
Email classification works best with the other processing steps: classification sorts the incoming stream, Scan & Recognise turns PDFs into structured data, and enrichment and booking prepare the invoice for your software. Together they form one chain from mailbox to accounting.
The Email Receiver analyses the attachments of every incoming email. PDF files are assessed for invoice content via IDR recognition. XML files are checked for a valid UBL format and Peppol compliance. Emails without recognised invoice content, such as quotes, reminders or loose correspondence, stay in the regular inbox and are not forwarded into processing.
Recognition improves as more invoices from the same supplier are processed, because the IDR builds supplier profiles. An invoice that has been processed correctly once is recognised faster and more accurately on a later email from the same supplier.
Not if you use the O365 integration. With a direct link to an existing Office 365 mailbox, the platform monitors the inbox without suppliers having to change anything. Invoices that arrive at the existing address are recognised and processed automatically.
If you use a dedicated eConnect receipt address, the organisation communicates that address to suppliers once. The address is stable and does not change with subscription changes or administrative adjustments. For new suppliers you can include the dedicated address in your own templates by default.
Each attachment is assessed separately. If an email contains three PDF attachments, the system analyses each file independently. Attachments recognised as an invoice go to the IDR for further processing. Attachments that are not an invoice, such as a packing slip or a cover letter, are not forwarded.
If the same email also contains an XML attachment that is valid as an e-invoice, that XML is processed directly through the regular processing flow. Invalid XML in the same email results in an error message to the contact email address, while the other attachments are still processed.
After activation, the platform monitors the designated Office 365 mailbox for new messages. Incoming emails are assessed immediately. Invoices are routed into processing; other messages stay available in the mailbox as usual.
Forwarding or an email-address change is not required. The integration is meant for organisations that may not or do not want to configure email forwarding internally because of security policy. The mailbox stays reachable for regular communication; the eConnect classification layer works in the background.
Email classification is the first step in the receipt phase. After classification, the IDR takes over the PDF invoice for recognition: supplier, invoice number, lines, VAT and references are read into structured data. The result goes to the platform as a draft invoice for workflow, approval and matching to open orders.
XML invoices skip the IDR phase and are processed directly as a validated document. Every invoice, whether it arrived via Peppol, email or a manual upload, then follows the same processing chain in the platform. There is no separate process per channel. See the full chain on AI processing or read more about Scan & Recognise.
One conversation is enough to decide which setup fits your mail volume, administration structure and existing infrastructure.
Schedule a meeting
Ask Cora