Customs brokers and freight forwarders can now have commercial invoices, bills of lading and packing lists read and interpreted automatically.
New Image Document Management capabilities announced by Descartes Systems Group use logistics-trained AI agents. Those agents are embedded directly in the company's customs and transportation products.
The agents read multiple document types associated with the same shipment. They interpret them in context, rather than digitising each document in isolation. The output is data ready to start a customs entry, produced without leaving the application they already work in.
Commercial invoice entry is the specific target. "Entering commercial invoice data is one of the most labour-intensive activities in the import process, especially when shipments include complex commercial invoices with hundreds of line items," said Scott Sangster, general manager, logistics service providers at Descartes.
"Rather than simply digitizing individual documents, our agentic solution interprets commercial invoices, bills of lading, packing lists and related shipment documents in context," Sangster said.
"It can extract, validate and prepare data that can immediately begin customs and transportation workflows with significantly fewer exceptions and manual corrections," he said.
The capabilities draw on the Descartes Global Logistics Network. Organisations process trade documents inside the Descartes applications they already use to create shipments or prepare customs entries.
The company argues this avoids the handoffs that standalone AI document processing tools require. Those tools bring separate applications, custom integrations and manual transfers between systems. Descartes describes the result as eliminating swivel-chair workflows.
What it claims to change
Descartes lists five outcomes. The first is less manual data entry, through AI-enabled extraction and interpretation of customs and logistics information from trade documents.
The second is faster customs entry and shipment creation, by preparing operationally ready data that can initiate transactions without repetitive rekeying.
Third is more consistent processing across related shipment documents, with fewer exceptions, manual corrections and rework in high-volume operations.
Fourth is the ability to scale operations more efficiently, with expanded document-processing capacity that supports growth without proportional increases in staffing.
The fifth is adoption without customer-trained AI models, complex integrations or separate applications, because the logistics-trained agents are already embedded in existing Descartes solutions.
The company is pitching the release against the appetite for yet another separate tool.
"While document processing remains one of the largest sources of manual effort in customs brokerage and logistics operations, most customers aren't looking for another separate AI application to manage documents," said Ken Wood, executive vice president of product management at Descartes.
"The goal is not simply to read a shipment document, but to quickly understand its logistics context and rapidly prepare the data needed to move the customs entry or shipment forward faster and more accurately," Wood said.
The Global Logistics Network connects shippers, carriers and logistics service providers. It covers last mile deliveries, domestic and international shipments, transportation rating and payment, global trade research, customs compliance and related regulatory processes.