The engine

Weeks of supply is a guess. This is the arithmetic.

Reorder points and quantities calculated per SKU from real demand variability, with safety stock sized to the service level you actually want. Lead times, MOQs, and container quantities applied to the timing — and the list sorted by what ships this week.

Suggested purchase orders ranked by when they are due

Safety stock sized to the service level you chose

95% and 99% need very different buffers, and the gap between them is real money. Planster sizes safety stock per SKU from its actual demand variability rather than applying one flat weeks-of-supply rule to a catalogue where nothing behaves the same.

Your constraints, in the timing

Lead times, minimum order quantities, and container quantities are applied to when and how much, not bolted on afterwards. Change a lead time from six weeks to nine and every order date that depends on it moves.

Inventory position counted honestly

On hand, plus on order, minus what is already committed. The number people usually plan against is the first of those three, which is how a brand with 40,000 units in a warehouse runs out.

Overstock flagged before it is dead cash

Slow movers and overstock get surfaced the same way shortfalls do. Carrying cost runs 20–35% a year, so the SKU quietly sitting on eleven months of cover is a decision too.

Overnight

What the agent does with reorder recommendations

Every night

The agent re-runs this every night and ranks what it finds by due date, so the top of your morning list is what has to be ordered this week rather than what is merely low.

What it can’t do

It cannot place an order, approve one, or send anything to a supplier. It brings you the list and the arithmetic behind it.

Four hundred rows, four decisions

The purchasing review used to be a 400-row export sorted by whatever seemed urgent, worked top-down until the meeting ended. Ranked by due date against real lead times, the list that actually needs a decision this week is four items long. The other 396 are fine, and Planster can say why.

20–35%annual carrying cost on inventory you did not need

See this week’s four decisions

Connect your data and get the ranked list against your own lead times and MOQs.