ISO 9001 with AI — the certifier arrives and the dossier is already built.

The problem

The pre-audit sprint is the trap ISO 9001 never meant to set.

The standard does not ask for a sprint. It asks for continuous evidence that the quality system works: that every planned control was done, that every non-conformity was handled, that every corrective action closed the loop. What actually happens on the plant floor is a different story:

1. Evidence lives in islands — a control sheet signed on paper, the SCADA log of line 3, the supplier email with the batch certificate, the spreadsheet on the R&D manager's desktop with the latest recipe. Each island is looked after by whoever generates it, but none of them is connected to the dossier the auditor asks for.
2. The quality team ends up acting as human glue — the week before the audit it rebuilds files by hand, chases overdue signatures, writes up non-conformities that had been open for three months without closing. The quality system runs in sprints, not at steady state.
3. And the worst part — between one audit and the next, the system sleeps. The plant floor keeps generating data, but the QMS does not hear about it until the next sprint. By the time the auditor finds a deviation, it had been there for months.

The sentence the quality director has said in the steering committee more than once: "we have ISO 9001, but between audits nobody knows whether the system really works". That sentence is not weakness — it is honesty about a structural problem that, until now, had no affordable solution.

How it fits the IRIS system

iLEAN does not add another QMS — it seals the gaps between your QMS and the reality of the plant.

The quality system does not fail for lack of a QMS — it fails because there is a gap between what happens on the floor and what reaches the QMS. iLEAN acts as the filler for that gap. It does not ask you to change the QMS, or the ERP, or the sheets your people already use. It sits on top and guarantees that the evidence arrives complete, linked to the right batch, workstation and customer, and signed.

Tracer captures the evidence at second zero. Writer turns it into an auditable dossier. The agent spots the holes in the system and flags them before the auditor does.

The iLEAN pieces applied to ISO 9001 with continuous audit:

• iLEAN Tracer (Connect + integrations) — captures every piece of evidence the moment it is generated: a control sheet photographed by the operator, the log of a legacy machine over OPC-UA, the supplier email with the raw-material certificate, the forgotten spreadsheet of the R&D manager. It links it to the batch and the workstation. If the plant loses the network, capture keeps running on Edge.
• iLEAN Writer — turns the captured evidence into an auditable dossier with the structure ISO 9001 asks for: control records, corrective actions, management reviews, process indicators. The dossier is built while reality is being generated, not the week before the audit.
• Quality agent — cross-checks the captured evidence against the documented plan of the system. If a planned control has not been done, it sees it and flags the responsible workstation. If a non-conformity has gone 30 days without a corrective action, it sees it and flags it. If a supplier has certificates about to expire, it sees it and flags it. The person decides. The line does not restart on its own.

See the full IRIS architecture →

Before and after

Pre-audit sprint vs. live evidence with iLEAN

Aspect	QMS + pre-audit sprint	With iLEAN Tracer + Writer + Agent
State of the dossier between audits	Partial, fragmented across islands	Live, complete, ready today
Capture of plant-floor evidence	Manual: paper sheet → batch digitization	At second zero, linked to workstation and batch
Pre-audit sprint for the quality team	2-4 weeks of concentrated work	Review of evidence already built — days
Detection of gaps in planned controls	The auditor finds them	The agent flags them first — the person decides
Non-conformities left open	Detected at month-end (or in the audit)	Detected the day after the deadline
Time to reconstruct the traceability of a batch	Days — searching island by island, calling shifts	Minutes — complete dossier, automatic

Impact estimate

Impact estimate for your plant — to be validated with your own numbers.

The block below is an estimate to be validated with the data from your QMS and your quality team. We lay it out so the steering committee has an order of magnitude; we refine it during the diagnostic.

• Mid-sized industrial plant with ISO 9001 + 1-2 vertical standards (IATF 16949 / ISO 13485 / ISO 22000), a QMS in place and a quality team of 3-8 people.
• First value expected within a few weeks: digitization of the plant-floor evidence of 1-2 critical processes (process controls + non-conformities) cross-checked against the QMS.
• Indicative payback between 4 and 9 months, depending on the monthly volume of control sheets and the cost of the pre-audit sprint in person-hours.
• Expected reduction in time spent on administrative evidence tasks: ≥30% (estimate to be validated). The hard lever is not the hours saved — it is that one major non-conformity avoided in the next audit pays for the whole pilot.

And the quality director's reasonable doubt

"What if the AI gets a control sheet or a certificate wrong?" — hallucination is a problem of free generation, not of anchored tasks. In tasks where the AI merely recontextualizes a piece of data (reading a control sheet and linking it to a batch, extracting the date from a certificate), the best models brought the error rate below 1.5% ^[1]. And even so, nothing critical is decided alone: iLEAN proposes, and the person signs. The system allows — and configures — any record relevant to the audit to pass mandatorily through human validation. The three safety rings exist precisely for this.

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