Automation that's live, working,
and paying for itself

The Problem

Stronghouse Logistics runs a busy delivery operation across West Michigan. Drivers are out in the field all day, texting photos of delivery slips to a shared group chat when a load is completed.

On the back end, that photo triggered a cascade of manual work. Office staff would scan the group chat, match each photo to the corresponding open order, download the photo, open Word or a PDF editor, manually create a delivery document, attach the photo, name the file correctly, upload it to the right folder in SharePoint, and then email the vendor to request payment.

Then they'd log the payment status in a spreadsheet. And reconcile everything at the end of the week. The owner was spending full days every Friday just making sure everything matched up. A missed photo, a wrong file name, or a forgotten email could cascade into payment delays and vendor friction.

The Solution

We built a fully automated delivery documentation system powered by AI. The key constraint: drivers couldn't change their behavior at all. The system had to accept an SMS photo (exactly what they were already sending) and handle everything from there.

When a driver texts a delivery photo, our automation intercepts it instantly. Claude AI reads the image, extracts the delivery information (date, load number, customer, etc.), and matches it against the open order database. A professional PDF is assembled with the photo, extracted data, and any required fields populated. The document is named correctly and filed automatically in the appropriate SharePoint folder.

Then the vendor payment request is drafted and sent automatically. No human in the loop. No manual steps. The whole sequence runs in under 30 seconds.

We also built a dispatch dashboard that gives the owner a real-time view of driver status, load assignments, and delivery confirmation. No more calling drivers or digging through spreadsheets.

The Result

The owner estimates they've recovered approximately 20 hours of admin time per week. Friday reconciliation went from a full-day grind to a quick review of the dashboard. Drivers report zero change to their workflow, because there was none.

The system has been running in production for months without a manual intervention. Zero delivery slips missed. Zero misfiled documents. Zero late vendor payment requests.