Skip to content
Case study · Logistics & Freight

Automatically validating shipping documents before reconciliation

We built an OCR solution that cross-checks supplier invoices, house bills of lading and carrier invoices, flagging quantity mismatches before they become costly reconciliation problems.

3

document types cross-checked automatically

The challenge

The finance team had to confirm that quantities matched across supplier invoices, house bills of lading (HBLs) and carrier invoices before reconciling each shipment. Their existing system couldn’t compare data across these documents, so the checks were done by hand. It was slow, and mismatches were easy to miss.

Our approach

We worked with finance and operations to define exactly what “matching” meant, including acceptable tolerances and how each document type is laid out. We were upfront that the existing system couldn’t do this natively, so we designed a solution that sits alongside it rather than waiting for a system change.

The solution

  • OCR using the Mistral AI API to extract quantities and key fields from each document
  • Matching logic that compares the three documents for every shipment
  • Azure SQL Database and Blob Storage to store extracted data and original documents
  • A simple web interface showing mismatches for the team to review

The outcome

The team now reviews only genuine exceptions instead of checking every document. Reconciliation is faster, human error has dropped, and recurring supplier issues are easier to spot and raise.

Ready to analyse, automate and accelerate?

Tell us what you're working on. We'll get back to you within one business day with ideas on how we can help.