The problem
In parcel delivery, every minute counts. Under time pressure, couriers often sign themselves instead of getting the recipient's signature, and simply leave the parcel at the door.
The system then shows a valid signature. Nobody notices until someone complains. Manual checks only cover a sample, and they take a lot of time.
- Parcels get lost, and investigations and replacements cost money
- Forged signatures are forgery of documents
- Parcels left in the open can be a data protection issue
Not an isolated case, but a pattern
How big is the problem really? In a pilot analysis we evaluated almost 100,000 signatures from three depots. The result: forgeries are not the slip of a few individuals, they show up on almost every route.
Almost a third of couriers forge more than 90% of signatures.
Three typical patterns
This is what a single route looks like after Signature CHECK has checked it. Each box is a signature that was flagged as forged.
Imitation: Different names every time, but clearly the same hand. 58 forgeries on one route.
Initials: A quick scribble instead of a real signature. 75 times on one route.
False safe drop: Parcel left “at the door” and signed off with initials.
From spot checks to checking every route
Capture
Every signature on a route is imported and analysed automatically.
Detect
Two specialised AI models find imitations, initials and recurring patterns.
Notify
Suspicious routes and couriers trigger a warning in the dashboard.
Act
Problems become visible before complaints come in. Important senders can also be monitored closely.
All routes at a glance
The dashboard shows how the forgery rate develops, which routes stand out and which couriers are involved. One click on a route shows every single signature, sorted into valid and forged.
Signature CHECK runs on your own servers or in a private cloud. Results can be exported, for example to Power BI or QlikSense.
Research that reaches everyday work
Signature CHECK is built on our own research. We developed and trained the models ourselves, together with Johannes Kepler University Linz and the University of Applied Sciences Upper Austria.
- Our own models for image comparison and pattern recognition
- Trained on more than 4 million examples labelled by people
- Runs on our own infrastructure, GDPR compliant
Happy to tell you more about our AI research and what it could do for your business.
What Signature CHECK delivers
Fewer losses, investigations and manual checks.
Spot legal and data protection risks early.
The team only checks what the AI flags as suspicious.
Higher delivery quality as a real edge over the competition.
You might also be interested in
HF Elektrik
Work reports on paper, times recorded twice, invoices copied by hand for days. Today the field technicians fill in their daily report on the tablet, even offline, and the times flow into time tracking automatically, including surcharges.
View project →
CargoCommander
Orders through various channels, paper and Excel, credit notes matched by hand. CargoCommander connects the ÖBB Rail Cargo Group, dispatch and drivers on one platform.
View project →
Teply ticket system
For Teply, a specialist in installations and service call-outs, we built a ticket system that handles customer requests, repair jobs and the coordination of technicians digitally, right through to simple invoicing.
View project →Any questions?
We are happy to tell you more about the project and what parts of it could work for you too.
