Lab Excel, into the ERP

The problem

The lab's real numbers live in a spreadsheet the ERP never opens.

Two parallel worlds — the real Excel with live shift data, and the ERP's quality module with data transcribed late by someone who spends part of their day copying and pasting, and makes mistakes.

• Every shift the quality manager records crumb moisture, average piece weight and specific volume per sample in an Excel sheet on a shared folder.
• The ERP's quality module holds a second version of the same numbers, typed days later by someone who spends part of the day copying and pasting between two windows. Two parallel worlds, and the official one is always the late one.
• Copying mistakes appear at the worst moment: a retail customer asks why a week of pan loaves ran light and the two worlds disagree on the answer.
• Meanwhile several lab people hold ERP licenses whose only use is entering results that already exist in the spreadsheet, a cost that renews every year without anyone questioning it.

How it fits the IRIS system

Connect in listening-inward mode — it reads the saved file, not the person.

Connect's "listening inward" mode — it watches the shared folder (not the person), reads the delta when the file is saved, parses it with a grounded language model, verifies it on screen, and pushes it to the ERP via API.

The lab keeps its spreadsheet exactly as it is, with its own tabs, colors and habits. What disappears is the second, late copy inside the ERP, and with it the licenses that only existed to type that copy. The quality manager stays in the loop as the one person who validates before the push.

See the full IRIS architecture →

Before and after

Lab spreadsheet with retyping versus lab spreadsheet connected

Aspect	Today	With iLEAN Connect
Gap between lab and ERP	Days	Zero latency
Hours spent retyping each month	Several	Zero
Crumb moisture and specific volume	Two versions that disagree	One verified record per batch
Average weight trend by line	Built by hand when asked	Already in the ERP
Changes to the spreadsheet layout	Break any macro	Absorbed as they come
ERP quality-module seats	One per person typing	One, for the manager who validates

Impact estimate

Impact estimate — to be validated against your ERP invoice.

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.

• 💰 CFO angle: the ERP bills per named user. If several lab staff log in only to enter quality data, only the validating manager keeps a seat: a typical 50-80% reduction in quality-module licenses.
• Estimated payback 4-9 months.
• From days of lag between lab and ERP down to zero, and hours of monthly transcription down to none, which go back to analysis instead of copy and paste: trends by line, by oven and by flour batch.
• And one version of crumb moisture and specific volume per batch, instead of two that have to be reconciled the day a retail customer questions a delivery.

If several lab staff access the ERP today purely to enter quality data, that's several module licenses. With Connect, only the validating manager needs one — a typical 50-80% reduction in quality-module licenses. Estimated payback of 4 to 9 months. Estimate to validate.

And the fair question from the production manager

“Our lab sheet has been edited by five people over the years — will it survive?” — that is the expected case: parsing does not depend on fixed cell positions, it reads columns, units, merged cells and notes in the margin as they are. It is an anchored task, where the best models drop below 1.5% error [1], and the quality manager verifies on screen before anything is pushed to the ERP.

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