The traceability of everything you will sell is born in a receiving desk that is typed in a hurry

The problem

The traceability of everything you will sell is born by typing in a hurry.

Receiving is the exact point where the traceability of everything the plant will later sell is born. And it is also the point where things get typed by hand, in a hurry, with the carrier charging demurrage. In a hurry, a heat number gets mistyped. That error does not show up today: it shows up a month later, in a claim that can no longer be traced because the chain was broken from minute one. On top of that, the mill test certificate arrives on its own schedule — paper or PDF — sits in an email, and has to be hunted down when the customer asks for it.

• The truck arrives with six coils. The warehouse lead has the paper packing list, the tag in the eye of each coil and a mill test certificate that sometimes turns up days later.
• They have to sign, book each coil into the ERP and get the crane moving, with the carrier charging demurrage. In a hurry, a heat number gets mistyped.
• That error does not show up today: it shows up a month later, in a claim that can no longer be traced because the chain was broken from minute one.
• And the mill test certificate — paper or PDF — sits in an email and has to be hunted down when the customer asks for it.

How it fits the IRIS system

Connect in photo mode with ERP write-back — packing list, tag and certificate, no mill changes.

Connect in photo-to-analog mode with ERP write-back. AI vision recognizes each mill's packing list layout without asking any of them to change it, and extracts the tabular structure. For coil tags: barcode, QR or plain text by OCR. The mill test certificate, paper or PDF, is anchored to its coil from minute one. One validation tap on the tablet and the coils enter the ERP with the correct heat.

Every mill has its own packing list layout and nobody is going to change it for you: the model recognizes the tabular structure of each one as it is. The tag is read by barcode, QR or plain text. And the certificate is anchored to its coil from minute one, so proving what you sold stops being a search through an inbox.

See the full IRIS architecture →

Before and after

Today's receiving versus captured receiving

Aspect	Today	With iLEAN Connect
Time per coil truck	45-60 minutes of typing	Under 10 minutes
Heat number errors	Invisible until the claim	Zero at source
Mill test certificate	In an email, hunted for later	Attached to the coil from minute one
Packing list layout	One per mill, typed by hand	Recognized as is, per mill
Carrier demurrage	Recurring	Reduced
Receiving and warehouse licenses	2-3 active users	One tap by the lead, no login

from 45-60 minutes of manual entry per truck to under 10. From heat number errors invisible until the claim, to zero errors at source. Certificate attached from the start instead of hunted for afterwards.

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.

• Estimated payback 3-8 months, plus reduced carrier demurrage.
• 💰 CFO angle: the receiving and warehouse module typically requires 2-3 active licenses; with Connect only the lead validates with one tap, without logging in. 50-70% reduction in receiving/warehouse module licenses, plus the mass capture module you no longer need to buy.
• From 45-60 minutes of manual entry per truck to under 10, truck after truck.
• And a heat-to-coil chain that is correct from minute one, which is what a claim a month later depends on.

CFO angle — the receiving/warehouse module typically requires 2-3 active licenses (warehouse lead, administrative support, supervisor). With Connect only the lead validates with one tap, without logging in: a 50-70% reduction in receiving/warehouse module licenses, plus avoiding the mass capture module usually sold separately. Estimated payback 3-8 months, plus reduced carrier demurrage. *An estimate to be validated*.

And the fair question from the production manager

“Does it recognize every mill's packing list?” — yes, and without asking any of them to change theirs: the model extracts the table structure with no need for a per-mill template. Tags are read by standard symbology where present and by optical recognition where not. It is an anchored task, where the best models drop below 1.5% error [1], and the packing list, the tag and the certificate are cross-checked against each other before the lead validates, so a heat that appears in one and not the others is flagged.

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