The batch exists in the system from second zero

The problem

The formula that was actually dosed lives in the margin of a printed sheet.

Every bioinput batch starts with a printed work order: which active ingredient, what dose of each raw material, in what sequence. Today that paper piles up in a folder until someone transcribes it, sometimes days later. With iLEAN Connect, the operator takes two photos with their phone: the work order and the first control label. Zero training, zero change in routine.

• Every batch starts with a printed work order: which active ingredient, which Bacillus or Trichoderma strain, what dose of each raw material and in what sequence it goes into the formulation tank.
• The real batch is rarely the printed one. A carrier topped up because a drum came in denser, an adjuvant added later, a pH corrected by hand: those adjustments are written in pen and nowhere else.
• The sheet then piles up in a folder until somebody transcribes it — hours later on a good day, days later in a week with several campaigns running.
• If a field complaint about efficacy arrives in that window, the batch physically exists, has already been bottled and may have shipped, but no system can say what went into it.

How it fits the IRIS system

Connect in photo mode — two pictures, and the operator's routine stays as it is.

A model grounded in this plant's reality reads both photos, extracts active ingredient, dose, raw-material batch and notes, and structures them automatically.

The operator does not fill in anything new: they photograph the work order and the first control label, the two documents they already handle. What changes is that the formula actually dosed — pen notes included — stops depending on someone typing it before it gets lost. It is usually the first case approved precisely because it asks the floor for almost nothing.

See the full IRIS architecture →

Before and after

The work order in a folder versus the work order captured

Aspect	Today, on paper	With iLEAN Connect
When the batch exists in the system	Hours or days after dosing	Second zero
Pen adjustments to the formula	Lost unless somebody retypes them	Extracted and tied to the batch
Strain lot used in the batch	On the sheet, if it was noted	Linked from the first photo
First control label (pH, temperature)	Filed with the work order	Read and stored as the batch baseline
Answering an efficacy complaint	Searching folders	One query by batch number
What the operator does differently in biosolutions	—	Two photos with their phone

Before: hours of latency until transcription, with formula adjustments scattered on paper. After: the batch exists in central memory from second zero, traceable from minute one. Estimated payback 3 to 7 months depending on daily batch volume.

Impact estimate

Impact estimate — to be validated with your daily batch count.

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 3-7 months, depending on how many formulation batches you start per day.
• From hours of transcription latency to a batch that exists in central memory from second zero, traceable from minute one.
• Formula adjustments stop being scattered across paper and become part of the batch record, which is what a registration holder needs to show.
• Transcription hours returned to production and quality, who today spend them copying sheets they have already read once.

Estimated payback 3-7 months · from hours of latency to zero-second batch start.

And the fair question from the production manager

“And if it misreads a dose or a strain lot written in pen?” — reading fields from a work order whose layout is always the same is an anchored task, where the best models drop below 1.5% error [1]. And there is a second net: the summary goes to the tablet at the line, and the operator confirms or corrects it before a single value crosses into the ERP.

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