ipp. easy13485 ipp. academy Field Guide · Vol. 03 00
Topic of the month · AI-only QMS risk No. 03 · 2026
ipp. academy
Expert inside. Robot supervised.

Don’t let the robot hold the clipboard.

A field guide for founders who asked AI for a QMS and got 47 SOPs, one existential crisis and a Notified Body email.

Turn the page
Chapter 01 · The struggle

The garage has a QMS. Sort of.

A founder asked AI to “make us ISO 13485 ready.” It delivered a folder called final_final_really_final and a CAPA template with motivational quotes.

That is the real comedy: not AI being bad. AI is useful. The risky part is treating generated text as if it were a controlled, scoped, role-owned, audit-defensible quality system.

I admire the ambition. I’m less excited about the orphaned risk file named “ask-chatbot-v6.docx”.
Template debt47 SOPs. Zero process ownership.
NB email“Please explain the rationale by Friday.”
CER folderClinical evidence? More like clinical vibes.
11:48 pmAudit prep begins. Espresso becomes a department.
From: Notified Body
Subject: Kindly provide traceability, rationale, signatures, current version, training proof and your soul.
Chapter 02 · The turn

AI is a power tool. It is not the PRRC.

Use AI to find, draft, compare, summarise and rehearse. Do not use AI as the sole guide for intended purpose, classification, risk acceptance, clinical strategy or QMS sign-off.

AI can draft a process.It cannot prove the process fits your product, scope, role model and lifecycle.
AI can sound confident.It cannot carry accountability when an auditor asks why that decision was made.
AI can summarise MDR language.It cannot replace expert judgement on claims, evidence gaps or clinical pathways.
AI can create a folder.It cannot make the folder controlled, trained, reviewed, maintained and defensible.
Good use

Search & orientation

Speed up navigation, drafts and first-pass understanding.

Danger zone

Decision by autocomplete

Never let plausible prose become your regulatory rationale.

The fix

Expert-gated workflow

AI in the system, experts at the decision points, evidence in the PLC.

Chapter 03 · The fix

The Makerspace is where the mess gets sorted.

The point is not “less expertise.” The point is putting expertise where it matters: claim logic, scope, product lifecycle, clinical evidence, risk, audit readiness and QMS operation.

easy13485 gives the startup a structured operating model: roles, guided processes, templates, registers, training and AI support. IPP experts step in as quality gates, not as a permanent bottleneck.

Makerspace
operating
model
01Claim & product line clarity
02PLC-based evidence build
03Roles, training & accountability
04Templates, registers & traceability
05Expert gates before audit pain
Chapter 04 · The centerfold

The Journey turns panic into a path.

The sequence matters. First clarify what the product claims. Then build the evidence and QMS around the actual route — not around generic template debt.

Chapter 05 · Capabilities in context

Not a service menu. A startup de-risking sequence.

Every capability answers a specific founder panic. Start with the uncertainty that hurts most; move forward without rebuilding the whole regulatory universe from scratch.

Question

“What are we actually claiming?”

Start with Claim Check before you write a QMS around the wrong intended use.

Open Claim Check
Operating model

“How do we stop improvising?”

The Makerspace combines guided QMS, Academy, documentation, AI support and expert gates.

Open Makerspace
Evidence

“Where does the technical file come from?”

Evidence grows along the product lifecycle instead of being reconstructed before review.

Open Evidence
QMS operation

“Who keeps this thing alive?”

QMSaaS is the maintained Easy13485 workspace for roles, processes and daily operation.

Open QMSaaS
Audit readiness

“What happens when the auditor walks in?”

Readiness audit and support turn weak spots into CAPAs before the Notified Body finds them.

Open Audit Support
Clinical

“Do we need a study, literature, PMCF — or a miracle?”

Clinical evaluation and study-design expertise connect claims to evidence, not wishful thinking.

Open Clinical Evaluation
Subtopic · Clinical evidence

Clinical study design is not a chatbot side quest.

AI can help structure literature and draft questions. Experts still need to decide what evidence actually supports the claim: equivalence, literature, performance route, PMCF, or a clinical investigation design.

This is where the ipp. clinical team matters: the claim, the patient/user context, the state of the art and the evidence route have to match. Otherwise the CER becomes a beautifully formatted argument nobody believes.

Ted’s clinical sanity check
ClaimWhat do you say the device does, for whom and in which use context?
RouteIs equivalence plausible, is literature enough, or do you need generated clinical data?
Evidence gapWhat is missing before a reviewer can believe the conclusion?
LifecycleHow does the evidence connect to risk, PMS and PMCF after launch?
If your clinical strategy starts with “the AI said probably,” I’m putting on the red glasses.
Chapter 06 · Straight from the site

The joke opens the door. The website carries the depth.

This brochure should tease, orient and make the founder smile. The detail lives in the capability pages.

Journey

From Claim Check to operated QMS

Use the Journey page as the main story page after the brochure.

Open Journey
Warranty

The quality warranty framing

For the serious proof that easy13485 is not just templates and vibes.

Open Warranty
Audit

Readiness before the audit week

The place to send people who still believe audit prep happens at 11 p.m.

Open Audit Support
ipp. academy Field Guide
Vol. 03 · AI-only QMS risk · 2026
Make the choice

Keep building your QMS forever — or start the easy13485 path.

The goal is not to remove expertise. It is to stop wasting expert time on chaos. Let AI accelerate the work, let easy13485 structure the work, and let ipp. experts check the decisions that matter.

StartClaim Check
Journey Calculator
BuildMakerspace
Evidence & QMS
ProveClinical evaluation
Audit support