Most tools look good in a demo with ten clean documents. A freight forwarder does not have ten clean documents. It has thousands a month, from hundreds of shippers, carriers and customs brokers, in several languages, and half of them arrive in the wrong email thread.

The right intelligent document processing for freight forwarders is one that handles that reality. It takes the volume without a person in the loop for every file, groups the documents of one shipment into one case, reads commercial invoices and customs declarations correctly, works in any language, connects directly to the ERP, and stops exceptions with a reason instead of letting bad data through. DOConvert was built for exactly this.

Here are the criteria to check before you choose, and how the flow works in practice.

The six criteria that matter for a forwarder

Volume without manual review per file

Thousands of documents a month means the system has to work unattended most of the time. Ask what share of documents needs a human touch, and what happens when a new shipper sends a layout the system has never seen. DOConvert has no template per sender, so a new shipper or carrier works on the first document.

Grouping documents into one shipment case

A shipment is a commercial invoice, a packing list, a bill of lading or air waybill, and often a certificate of origin and a customs declaration. They arrive separately. The system must recognise they belong together, by booking number, container, BL number or invoice reference, and hold the case until it is complete.

Commercial invoices and customs declarations

These are the documents where mistakes cost money. HS codes, country of origin, Incoterms, currency, values per line. The system has to read them as structured line items, not as a block of text, and check them against the shipment record.

Multi-language

Invoices from China, Turkey, Germany and Brazil in the same mailbox. The system reads each document in its own language, without a separate setup per language.

Direct connection to the ERP or freight system

The data has to land in the shipment or job record, with the right references. Ask whether the connection is direct and bi-directional, so the system can read the booking before it writes to it. DOConvert connects directly to Priority through the DOConvert Priority connector and to Oracle Fusion Cloud ERP, no middleware in between, and the data does not leave your environment.

Exceptions stopped with a reason

A value that does not match the booking, a missing HS code, a packing list with fewer cartons than invoiced. The system should hold the case and say why, in plain words, so an operator fixes one thing instead of re-checking everything.

What goes wrong with the wrong tool

OCR that reads but does not understand

Text comes out, but the quantity ends up in the price column and the total is read from the wrong page. Someone still checks every file, so the volume problem is not solved.

Templates per sender

With hundreds of shippers, a template per sender means a setup queue that never ends. Every new customer waits for a template.

Data entered without a check

The system writes to the ERP whatever it read. Errors are found by customs, by the customer, or by finance weeks later.

How it works with DOConvert

One shipment, one case, one entry

DOConvert is the AI layer for the ERP. Here is the flow for an import shipment.

• The documents arrive by email, often across several messages and days
• DOConvert reads each one, works out what it is, and groups the commercial invoice, packing list, bill of lading and customs documents as one shipment case
• It identifies the shipper, the consignee and the forwarder’s own job reference
• It finds the open shipment or import record in the ERP
• It checks invoice lines, values and currency against the record, and packing list quantities and cartons against the expected goods
• Anything that does not match is stopped with a plain reason
• Only after the case matches, DOConvert enters it through the bi-directional connector

What the operations team sees

Operators see a short list of exceptions with reasons and the documents next to them. The rest of the shipments are in the system before anyone opened the email. A typical implementation takes about 10 business days, cloud or on your own servers.

For how this holds up at volume, see logistics document processing at high volume.

For the commercial invoice flow in detail, read how to automate commercial invoices in logistics. On why a model built for supply-chain documents reads them better than a general one, see Generalist vs specialist, the real debate in IDP models.

Who it fits

• Freight forwarders and customs brokers handling thousands of shipment documents a month
• VP Operations who want shipments in the system before the vessel arrives
• Finance teams at forwarders that reconcile supplier and carrier invoices against jobs
• IT managers who want a direct ERP connection and no middleware to maintain

Questions and answers

How does the system know which documents belong to the same shipment

By the references on the documents, such as the booking number, container number, BL or AWB number and invoice number, and by the shipper and consignee. DOConvert holds the case open until the expected documents are in.

Does it need a template for each shipper or carrier

No. DOConvert reads the document without a template, so a new shipper works on the first file.

Can it read documents in several languages

Yes. Each document is read in its own language, with no separate setup per language.

What happens with a document that does not match the booking

The case is stopped and the reason is written in plain words. Nothing is entered until an operator fixes or approves it.

Which systems does it connect to

Priority through the DOConvert Priority connector, Oracle Fusion Cloud ERP and Infor M3, all directly. SAP is in progress.

Want to see it on your own documents? Book a Demo or Get Started