Stock transfer planner
A small system that spots stock sitting in the wrong branch, checks whether moving it is actually worth the cost, protects the site it would come from, and waits for a vehicle that is going there anyway. It also handles the thing most stock systems get wrong: goods that have left one place and not arrived at the other. Seven posts on the same system, one diagram at a time, with a cost breakdown and an engineering reference at the end.
- 01
A stock transfer planner on AWS for a few dollars a month
The whole system on one page — imbalance, worth it, batch — and the test that removes most of the proposed transfers before anybody moves anything.
- 02
When a transfer is actually worth it
What a transfer actually costs at both ends, how the benefit should be computed, why low value items rarely qualify, and how ping-pong transfers are prevented.
- 03
How the source gets protected
Why the source needs its own cover rule, the difference between surplus and display stock, how transfers are sized, and the site that is always the donor.
- 04
How transfers get batched onto movements that already happen
Why batching changes the economics, how long a proposal can wait, how a full movement is assembled, and when a dedicated trip is justified.
- 05
What in-transit stock does to your numbers
Why in-transit needs to be a real state, the two failure modes of not modelling it, how transfers are confirmed at both ends, and the stock that never arrives.
- 06
What the stock transfer planner costs
About $122 a month. One Bedrock read per SKU-site pair is the only line that grows; the queue, the table, the mail and the storage are rounding errors. Plus the three ways the bill could surprise you.
- 07
Engineering reference: the stock transfer planner architecture
Same system, drawn purely for engineers. Service names, region, Lambda inventory, IAM scopes, the schemas and the exact model id.
Frequently asked questions
- What is a stock transfer planner?
- A small serverless system that identifies stock imbalances between sites, tests whether a transfer is economically worthwhile, checks it will not cause a shortage at the source, and proposes it on an existing vehicle movement where possible.
- Why are most transfers not worth it?
- Because the cost of picking, packing, moving and receiving is often larger than the margin on the units being moved, particularly for low-value items.
- What is a ping-pong transfer?
- Stock moved from A to B and then back again a few weeks later, usually because both transfers were triggered by a low stock level rather than by real demand.
- Why does in-transit stock matter?
- Because it is simultaneously not at the source and not yet at the destination, and systems that do not model it either lose it or count it twice.
- What does it cost to run?
- A couple of dollars a month. See part six.