Calderwood Freight

Calderwood was never short of software. It was short of a place where the decision lived. Route changes arrived by phone, capacity sat in a spreadsheet nobody owned, and the planner who knew why last Thursday went sideways was usually on leave.

So we did not add a tool. We moved the decision itself. One service now reads the shipment feed, weighs capacity against the promise made to the customer, and commits a plan. Judgement did not disappear; it started happening against a record rather than a memory.

Four things had to be true before any of it shipped:

• every input arrives on a feed, never in a forwarded email
• the rule that picks a route is written down, not remembered
• anything done the same way twice belongs to the machine
• a planner can always see why a plan changed, and override it

Consistency now comes from the logic rather than from whoever is on shift. Two planners facing the same board produce the same plan, which took most of the argument out of Monday morning.

Nothing Calderwood already owned was thrown away. The service reaches into the existing transport platform, the carrier feeds and the finance export, so nobody had to learn a second way of working to see the benefit.

The week changed shape after about a month. Planning stopped being a morning of firefighting and turned into a review of what had already been proposed overnight. When volume climbed in the fourth quarter, the extra load landed on the service instead of on overtime.

Calderwood did not buy automation. It moved where a decision gets made, and everything downstream followed.