Scale and WMS, stitched to the ERP with no integration project

The problem

Three systems that could talk to each other, and a person in the middle retyping.

The cutting scale and warehouse WMS have their own APIs, but nobody's had time to properly integrate them with the commercial ERP. Every weighing or location move gets keyed into the ERP by hand again.

• The scale weighs, the WMS knows where every pallet of formats is, and the ERP is what invoices the customer. All three are modern enough to have an API of their own, and none of them has ever been introduced to the others.
• Nobody has ever had the two or three weeks it would take to integrate them properly, and that project has sat at the bottom of the list for years.
• So every weighing and every location move gets keyed into the ERP by hand, a second time, and in a business that invoices by the kilo a mistyped weight is money.
• The result is the recurring argument between what was invoiced and what was delivered, which always ends up being settled by weighing it again.

How it fits the IRIS system

Connect stitches what already exists — no scale, WMS or ERP is replaced.

Connect stitches the systems together via the API that already exists, without replacing the scale, WMS or ERP. Weight, size and location reach the ERP in the background.

This is not an integration project, and that is the whole point: it does not compete for the same weeks, it opens nothing and it swaps nothing out. The weight the scale gives is the weight the ERP invoices, and it gets there on its own.

See the full IRIS architecture →

Before and after

The isolated scale versus the stitched scale

Aspect	Today	With iLEAN Connect
Weight of a cut order	Typed into the ERP by hand	Arrives on its own, from the scale
Location moves	Keyed in a second time	Synced in the background
Invoiced versus delivered	A recurring argument	The same number in both
A remnant back on the rack	Often never recorded	Reweighed and back in stock
Integration project	Years at the bottom of the list	Not needed
Scale, WMS and ERP	—	None of them replaced

Double data entry on every movement → automatic sync with no manual step.

Impact estimate

Impact estimate — to be validated with your numbers.

The block below is an estimate to be validated against your plant's actual data. We put it forward so the committee has an order of magnitude; we refine it during the assessment.

• Administrative time recovered on every movement, which in a warehouse that weighs all day is the bulk of the office's workload.
• Fewer invoiced-versus-delivered mismatches, which today are settled by reweighing and by issuing a credit note.
• The remnant that goes back on the rack gets recorded with its real weight instead of quietly disappearing from stock.
• And the integration project pending for years stops being a prerequisite for anything else.

Administrative time recovered and fewer invoiced-vs-delivered mismatches. Estimate to be validated.

And the fair question from the production manager

“If we are going to connect systems, isn't a proper integration cleaner?” — it is, on the day it reaches the top of the list, and that is exactly the problem: in years it never has. This stitching leans on the interfaces that are already documented, replaces nothing and asks for no shutdown window, so it does not join that queue. Matching a weighing to the right order line is an anchored task where the best models drop below 1.5% error [1], and any mismatch is raised for a person rather than written.

^[1] OpenAI paper "Why Language Models Hallucinate", 2025 — on the reliability of AI in anchored tasks.