Work · Logistics and distribution

You can see what every load earned, and none of them closes with weight missing.

Loads were chased on WhatsApp all day. Weights were argued about after the fact, signed delivery notes came back days later, and nobody could say what a trip had actually earned until somebody rebuilt it from a spreadsheet.

Client
A distributor running its own trucks

What we built

A load system running from the weekly demand plan to a closed, reconciled delivery. Each stage refuses to advance without what it needs: no allocation without an agreed rate, no dispatch without a gate pass and a delivery order, no closure until the transporter invoice or the fuel cost is in and the load carries no flags. Every advance records who did it and when. The weighbridge screen shows tare and gross and will not let anyone type the net. Drivers report by WhatsApp, and a driver who estimates a weight is asked for the weighbridge figure instead.

The load board, with the weight variance that closes a load or holds it.
A drawing of the interface, not a capture. The load board, with the weight variance that closes a load or holds it.

What happened once it was running

Margin per truck and per region is visible while the work is still happening rather than at month end. A load short beyond tolerance carries a flag that says what is wrong in plain English, and it does not close.

2.5%Weight tolerance before a load is flagged
6hBefore a loaded truck still on site is raised
3Days from first conversation to a working system
0Loads that can close with an unexplained shortfall
From the delivered system

The tolerance and every stage gate are configuration in the delivered system, readable rather than described. The stall clocks are set per stage rather than uniformly, because a signed delivery note legitimately takes days to come back and a loaded truck sitting in the yard does not. The three days is measured from the first message received to the deployed build.

Why this page is short

Every figure on this page traces to something we measured.

This page once carried claims that described the work inaccurately. They came down in an audit of our own copy, and each one went back up only when its numbers could be traced to something measured. A case study you cannot check is worth nothing to you, and one that turns out to be wrong costs us more than it ever earned.