One shipment produces a small stack of paper before it reaches the warehouse. A bill of lading from the carrier. An invoice and a packing list from the shipper. An arrival notice, a customs declaration, then the invoices for the freight itself. Freight document processing is the work of turning that stack into one checked record.
In most operations a person still does it. They open each PDF, find the shipment, retype the references and compare the numbers by eye. On a quiet day it is tedious. In peak season it is the reason files close late.
This article goes through the documents one at a time, what each is for and where it goes wrong. For the buying question, which platform to choose, see intelligent document processing for freight forwarders.
The documents, one by one
Bill of lading or air waybill
From the carrier or the forwarder. Shipper, consignee, ports, container or flight, dates. It ties every other document to one physical shipment, so a typo here spreads everywhere.
Commercial invoice
From the shipper. Items, quantities, unit prices, currency and terms of sale. Customs uses it to assess duty, and the buyer checks it against the purchase order.
Packing list
Also from the shipper, often sent separately. Cartons, weights and quantities per carton. The warehouse receives against it, which is why many teams process packing lists automatically before the container is opened.
Arrival notice
From the carrier or agent at destination. Arrival date, charges due and where to collect. It starts the clock on storage, and it is the document most often read too late.
Customs declaration
From the broker, after clearance. Classification, value, duties and taxes. It closes the customs side of the file and feeds the landed cost.
Freight and broker invoices
From the carrier, the forwarder and the broker. Freight, handling, storage and fees. Each has to be checked against the shipment it claims to belong to.
Certificates
Origin, conformance, inspection. Small documents that hold up a release when they are missing or do not match the invoice.
Why freight documents defeat most tools
They arrive separately
Different senders, different days, different mailboxes. A tool that handles one document at a time never learns that they belong together.
Every sender has its own layout
Each carrier, shipper and broker designs their own forms. Template-based tools need one template per sender and break when a layout changes.
One PDF can hold five documents
An agent scans the whole file into a single attachment. Before anything can be read, someone has to work out where one document ends and the next begins.
The value is in the cross-check
Does the packing list agree with the invoice? Does the invoice agree with the PO? Does the freight invoice belong to this shipment? Reading fields is the easy part. That cross-check is what makes an import file hard to automate.
What good freight document processing looks like
Documents are grouped by shipment
By references, parties and purchase order. The stack becomes one case, even when the documents arrive days apart.
Every document is checked against the record
The purchase order, the shipment and the import file in the ERP. Not just read, checked. The same holds for the customs documents the broker sends after clearance.
Exceptions stop with a reason
A weight that does not add up or a charge for the wrong container stops with a plain note. The team works on those and nothing else.
The connection to the ERP is direct
No middleware carrying shipment data through a third system. At high volume, every extra hop is another place for a file to stall.
How it works with DOConvert
DOConvert is the AI layer for the ERP. It connects directly to Priority through the DOConvert Priority connector, to Oracle Fusion Cloud ERP and to Infor M3, bi-directionally and with no middleware. Here is one inbound shipment.
The documents arrive by email
The invoice and packing list from the shipper, the bill of lading from the forwarder, the declaration from the broker. Each in its own email and its own format.
DOConvert identifies each sender
From the email address and the document. No template per carrier, shipper or broker.
A combined PDF is split
When one attachment holds several documents, each is recognised for what it is and handled on its own. That step is described in automated shipment document processing.
The documents are grouped into one file
By shipment references and purchase order. The purchase order and the open import file are read from the ERP.
The cross-check runs
Invoice against PO, packing list against invoice, declaration against shipment. A mismatch stops with a short reason.
The file is entered
Only when everything agrees, the documents and costs are entered against the import file and its PO. Nobody retypes a reference.
The shift from keeping paper to documents that are checked on arrival is the subject of The future of work, from paper trails to document evolution.
Who it fits
- Freight forwarders and customs brokers handling thousands of documents a month
- Importers and distributors with several inbound shipments a week
- VPs of Operations and controllers who need landed cost in the ERP without reconciling separate entries
- IT and IS managers on Priority ERP, Oracle Fusion Cloud ERP or Infor M3 who want a direct connection
Questions and answers
Which freight documents does this cover?
The documents that make up a shipment file. Commercial invoices, packing lists, bills of lading, customs documents and the invoices from forwarders and brokers. They are handled as one file, not as loose documents.
Do carriers and shippers need to change anything?
No. They keep sending the same documents to the same address. There is no portal and no agreed format.
What happens when one document is missing?
The documents that arrived are grouped, and the file waits with a reason that names what is missing. When the document comes in, it joins the file.
Is freight document processing the same as OCR?
No. OCR turns a page into text. Processing means the document is understood, tied to its shipment and checked before it is entered.
How long does implementation take?
A typical DOConvert implementation takes about 10 business days, in the cloud or on your own servers.
Want to see it on your own documents? Book a Demo or Get Started