The method

Six stages. What each one has to satisfy is stated openly; how it is solved is ours.

The order of these matters more than any of them individually. Most engines start at stage three and stop at stage four.

01.

Resolve

Establish what is physically on the order, at order drop, before anything is picked — dimensions at each packaging level, weight, shape class, orientation rules, bearing tolerance, compatibility. And how far each value can be trusted.

Outdimensions at every packaging level, weight, shape class · and a confidence on each

02.

Form

For goods whose shape is negotiable, decide the geometry. No other cubing engine treats product form as an output.

Outthe geometry the product ships in

03.

Compose

Carton, pallet and vehicle solved as one nested problem rather than three — because the best carton is frequently not the best pallet, and the pallet is what you actually pay to ship.

Outcarton, pallet and load solved as one problem

04.

Choose

Selected for downstream consequence, not immediate fit. The container that holds the most units is routinely not the container that ships cheapest.

Outthe container that ships cheapest, not the one that holds most

05.

Declare

Turn the physical result into the commitments that lock first — density and freight class, pallet and carton counts for the appointment and the ASN, packaging quantities for procurement, work content for the shift plan.

Outdensity and class · pallet and carton counts · packaging quantities · work content

06.

Reconcile

Compare against what actually shipped, attribute the difference, and feed it back. Confidence rises where confirmed and falls where it isn't — which is the only asset here that compounds.

Outconfidence that rises where confirmed and falls where it is not

Governing principles

The chain, beyond the box. The cheapest carton is rarely the cheapest shipment.
Early, and always fast. Knowing before the decision locks is where the value is. Solving in seconds once you ask is what makes it usable — we do both.
Measured, never assumed. Every saving we report is the difference between two real runs of the engine. Never an applied percentage.

Two entry points

Inbound Cube the order from scratch. Shape-native packing across cuboid, cylinder, sphere and conformable goods, under orientation, bearing, crush and weight constraints.
Outbound Ship whole cases from inventory first, then pool and repack only the residual — with case-breaking and ship-complete hold logic. Operationally correct, and not something a generic cartonization API does.