Lab spreadsheet to ERP in seconds

The problem

The lab's real numbers live in a file the ERP never sees.

two parallel worlds — the real spreadsheet, where the live data is, and the official ERP, transcribed weeks later by someone spending half a day a month copy-pasting and introducing errors.

• Every batch, the quality lab records pH, Brix, viscosity and salt in a spreadsheet: the readings that say whether the acidification holds and whether the emulsion is in spec.
• The spreadsheet is where the live data is. The ERP is where it is supposed to be, and it gets there weeks later.
• Someone spends half a day a month copy-pasting between the two, and a pH of 4.32 typed as 4.23 is exactly the kind of slip nobody catches.
• Meanwhile several lab staff hold ERP seats only to enter quality data, which is the licensing line finance asks about at every renewal.

How it fits the IRIS system

Connect listening inward — it watches the shared folder, not the technician.

Connect, "listening inward" mode — it watches the shared folder, not the person. It reads the delta the moment the file is saved, parses it with a grounded model, shows it for verification, and pushes it to the ERP via API once the manager confirms.

The lab keeps its spreadsheet exactly as it is. iLEAN reads the change the moment the file is saved, shows it for verification and pushes it to the ERP through its API. The two worlds stop being two. The technician keeps measuring and saving; the copy-paste day at the end of the month simply disappears.

See the full IRIS architecture →

Before and after

The lab spreadsheet today versus the spreadsheet connected

Aspect	Today	With iLEAN Connect
Lag between lab reading and ERP	Weeks	Zero seconds
Monthly copy-paste on sauce and canning lines	Half a day	None
pH and acidification per batch	In a file nobody queries	In the ERP against the batch
Viscosity of each emulsion	A column in a spreadsheet	A trend per formula and line
Transposed digits	Found by chance, if ever	Gone: nothing is retyped
Lab seats in the ERP quality module	One per technician	Only the person who verifies

weeks of lag between real data and ERP data → zero seconds. Half a day a month transcribing → zero.

Impact estimate

Impact estimate — to be validated against your real ERP license pricing.

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 4-9 months.
• CFO angle: if several lab staff log into the ERP only to enter quality data, with Connect only the person who verifies needs a seat, a typical 50-80% cut in quality and lab module seats.
• The lag between the real reading and the ERP goes from weeks to zero seconds, and half a day a month of transcription goes to zero.
• And pH, Brix and viscosity become queryable per batch the day a customer asks for them.

estimated payback 4-9 months. CFO angle: if several lab staff today log into the ERP just to enter quality data, with Connect only the person who verifies needs a seat — a typical 50-80% cut in the quality/lab module. *Estimate to validate against real vendor pricing.*

And the fair question from the production manager

"What if it misreads a column and sends a wrong pH?" — parsing a spreadsheet whose columns you defined is an anchored task, where the best models stay under 1.5% error [1]. The change still shows up for the quality manager to confirm before it goes to the ERP, and a value outside the physical range for that product, such as a pH no acidified sauce could have, is flagged rather than pushed. Nobody in the lab has to open the ERP to make that happen.

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