Home / Case studies / Shipping document validation with custom AI extraction
Al Dahra · Agribusiness & trading · Logistics & trade documents
Shipping document validation with custom AI extraction
Shipping-line documents validated against ERP data with custom-trained AI models and zero false positives. A shipping document validation automation project for Al Dahra, built with UiPath · Microsoft AI document extraction (custom models) · ERP integration.
The problem
Each shipping line sends shipment and invoice documents in its own format. Validating container, weight and amount data before paying logistics invoices took 15–25 minutes per shipment, done by hand.
What we built
UiPath identifies the shipments to verify using business rules, downloads the documents and works with the ERP. Custom AI models trained on Microsoft's AI document extraction platform extract and classify the data for comparison with the ERP baseline, with fail-safes on every match.
Results
- Validation fully automated
- 0% false positives, enforced by fail-safes
- Manual effort cut by more than 95%
How we'd build it today
An agent now handles new shipping-line formats without retraining, judges which mismatches are material, and drafts the query to the freight forwarder for a person to approve.
Built with: UiPath · Microsoft AI document extraction (custom models) · ERP integration
Related
More in Logistics & trade documents
Vessel schedule extraction into the TMS
15+ hrs/week Hours saved on schedule checksCarrier sailing schedules collected from shipping-line sites and imported into the TMS.
Bill of lading container count validation
< 30 sec Validation time per BL (from 5–10 min)Container counts on each bill of lading checked against port authority records before invoicing.
Tell us which task you'd switch off first
Bring one process. In 30 minutes we'll tell you whether it's worth automating, which KPI we'd move and how we'd get there in 8 weeks.