Flutter app · on-device OCR
DriverHub
A Flutter app for truck drivers: find and accept loads, scan paperwork into clean PDFs, and see what they'll be paid.
Working app, connected to the real API
- When
- Sep – Oct 2026
- My part
- Mobile developer: all app commits
- Credit
- The backend belongs to the client's team. Screens are from the Flutter app running on the iOS simulator with its demo data, company names made generic.
- Code
- ~12.2k lines of Dart
- OCR
- On device
- Data shown
- Demo only
Chapter 1 of 4
The right load, matched to the driver's lanes
Drivers see loads that match the lanes they like to run, with weight, distance, pay and the equipment they'll need, and accept one in a tap.
Chapter 2 of 4 ·
A phone camera that behaves like an office scanner
Drivers photograph bills of lading and delivery receipts in a dark cab. The app finds the page, crops it, cleans it up and names the PDF for them.
$ git log feat: Enhance document scanning and PDF generation features
Chapter 3 of 4 ·
Separating ink from paper
A small image-processing package of my own splits a document photo into paper and ink, then redraws the ink on clean white. Text recognition runs on the phone with ML Kit, so it works with no signal.
$ git log feat: Add ink_scan package for extracting ink from documents and enhance image processing capabilities
Chapter 4 of 4 ·
Wired to the real thing
OTP sign-in and the loads API are connected to the client's staging and production servers. The rest of the app runs on demo repositories behind the same interfaces, ready to switch over.