CASE STUDY · LOGISTICS
Proof-of-Delivery Paperwork, Filed Automatically for a Canadian Trucking Carrier
Drivers photograph the signed paperwork. Our AI reads it, finds the right order and files it, and sends anything uncertain to a person.
PODs a year on track to be filed automatically
hours of filing time a year on track to be saved
of PODs filed with no one touching them
filed on the wrong order
from driver photo to the dispatch system
The Problem
Every photo had to be filed by hand
Drivers photograph signed paperwork in the Samsara app. Office staff then matched each photo to the right order and stop in FleetManager, their dispatch system (TMS), and filed it by hand. Each one took about 8 minutes.
Multi-stop loads and paperwork numbers reused across orders made mistakes easy.
What We Built
An AI that files each photo on the right order
- 1Reads each photo and decides whether it is a proof of delivery.
- 2Finds the order using the numbers printed on the paper: BOL, PO and reference numbers.
- 3Confirms the consignee, date and driver against the dispatch system.
- 4Picks the right stop on multi-stop loads.
- 5Files the document into the dispatch system automatically.
- 6Feeds a custom ROI dashboard we built for the carrier's leadership, tracking PODs filed automatically and hours saved.
How It Stays Accurate
When in doubt, it asks instead of guessing
Uncertain means held, not guessed
Anything the system is not sure about is held for the carrier's team to answer with one click on a review page.
Answered papers are filed about 10 minutes later
Once the team answers, the paper is filed on the order they confirmed.
Held papers are re-checked
During the day, held papers are checked again as the dispatch system catches up.
The team stays in control
The carrier's team adds drivers themselves, and the team's answers feed new rules.
Tools It Connects To
Built around the carrier's existing tools
Samsara
Drivers photograph signed paperwork in the Samsara app. That is where each document starts.
FleetManager (their TMS)
Orders, stops, consignee, date and driver are checked here, and the finished document is filed here.
A review page
Uncertain papers wait here for the carrier's team to answer with one click.
Timeline
From pilot to production
Oct 26, 2026
Two-week pilot begins
The system starts reading and filing real driver photos.
Nov 10, 2026
In production
The system runs in production.
In the first three weeks, 280 PODs came in and 227 were filed automatically. That saved about 30 hours of filing, at about 8 minutes of staff time to file one POD by hand.
Common Questions
Questions people ask
How does the system know which order a proof of delivery belongs to?
It reads the numbers printed on the paper, such as the BOL, PO and reference numbers, and finds the matching order in FleetManager, the carrier's dispatch system (TMS). It then confirms the consignee, date and driver against that order, and on multi-stop loads it picks the right stop before anything is filed.
What happens when the system is not sure about a document?
It does not guess. Anything uncertain is held for the carrier's team, who answer with one click on a review page. Answered papers are filed about 10 minutes later, and held papers are re-checked during the day as the dispatch system catches up.
How much filing time does this save?
In the first three weeks, 280 PODs came in and 227 were filed automatically, saving about 30 hours of filing. Staff take about 8 minutes to file one POD by hand. The system is on track to save more than 5,000 hours of filing time a year.
Can the carrier's team add drivers without us?
Yes. The carrier's team adds new drivers themselves, without waiting on us or a developer. The answers the team gives on the review page also feed new rules, which is how the system keeps getting better at the carrier's paperwork over time.
Ready to see what AI can do for your business?
Book a free 30-minute consultation. We'll identify your highest-impact opportunity and outline a plan. No commitment required.
Book a free 30-minute consultation