Rebuilding Dispatch for a Regional Freight Operator
A fifteen-year-old scheduling tool was the reason every truck left late. We replaced it in twenty-seven weeks without stopping a single shift.
One Spreadsheet, Forty Dispatchers, No Undo
Dispatch ran on a desktop application written in 2009 and a shared spreadsheet that patched everything the application couldn't do. Two dispatchers editing the same route at once was a daily event, and the only fix was a phone call.
The vendor had stopped shipping updates. Drivers received their manifests as printed sheets, and any change after 6am meant a second print run or a radio call. The operator wanted a system they owned, on a timeline that didn't require a freeze on operations.
A Live Board and a Driver App That Agree
A single dispatch board that every dispatcher sees in real time, a driver app that replaces the printed manifest, and an integration layer that kept the old system fed until the last depot switched over.
Twenty-Seven Weeks, One Depot at a Time
A legacy migration is mostly reading. AI tooling mapped the 2009 codebase and the spreadsheet macros, generated the integration tests against recorded data and drafted the runbooks, so twenty-seven weeks went into decisions rather than archaeology.
Ride along
Two engineers sat with dispatch for three weeks of morning shifts before writing a line of code.
Board first
The dispatch board shipped to one depot in read-only mode, mirroring the old system, so trust came before control.
Drivers on
The app rolled out one route at a time, with printed sheets running in parallel until each driver asked to stop them.
Handover
Legacy system retired, runbooks and tests handed to the two engineers we helped the operator hire.
Have a System That Needs Replacing Without a Freeze?
Tell us what it runs on, and an expert team member will reply.