A notified body finding is rarely about a missing document. It is far more often about two documents that say different things. Cross-checking is the part a model does well.
MDR technical documentation is the evidence a manufacturer holds to show a device meets the EU Medical Device Regulation (EU) 2017/745. Annex II sets its content: device description and intended purpose, labelling and IFU, design and manufacturing information, the general safety and performance requirements (GSPR) checklist, benefit-risk analysis and risk management, and product verification and validation including clinical evaluation. Annex III covers the post-market surveillance documentation.
AI is useful here in three ways. Drafting sections from the controlled records that already exist. Mapping each GSPR to the evidence that demonstrates it, with the document, version and section cited. And checking consistency: that the intended purpose reads the same in the IFU, the clinical evaluation and the risk file, that every risk control is verified, that the harmonised standards cited are the current versions.
eQMS, DHF, document control
Annex II sections, GSPR links
IFU · risk · clinical · labelling
in your document control
versioned, with citations
two records disagree
Annex II of Regulation 2017/745 lists it: device description and specification, intended purpose and variants, labelling and instructions for use, design and manufacturing information, the general safety and performance requirements with evidence, benefit-risk analysis and risk management, and verification and validation including clinical evaluation. Annex III adds the post-market surveillance plan and reports. Most manufacturers organise it as a set of controlled documents.
It can check that each Annex II element is present, that each applicable GSPR has evidence mapped to it, and that references point to current versions and standards. That is a useful pre-submission check. It cannot judge whether evidence is adequate, which is the question a notified body actually asks, so the output is a list of gaps and conflicts for your regulatory team.
In your environment. The system reads from your eQMS and document control, writes drafts back as controlled documents, and runs on infrastructure you choose with no training on your data. Design history is commercially sensitive and complaint data can include patient information, so we design for that from the start.
What we built, what stage it is at, and a link to the live thing wherever there is one.
yourOS is an operating system with an AI at its heart. You talk to it like a person, out loud or typed. It makes your calls, plans your days and works on your projects with a crew of agents, on your own private server. Our own product, built from the first commit; the waitlist is open.
Website, backend and app assets for Sonora, a free AI sound-therapy app on iOS and Android published by Cognitive AI.
Brand, website, prototype, apps on five platforms, the AI backend and the algorithm-intelligence pipeline for Podswap, a social growth engine that runs without a human in the loop.
In closed beta: Roxy, a private AI assistant with permanent memory running on infrastructure the user owns. Brand, site, product, backend and conversational design.
Live at coherence.com: brand identity generated with AI agents we directed, the motion system and website for the group behind Sonora, AboutAI, WakeUp and MudShadows. Agent work remains in build.
Live at domainbooth.com for DomainBooth, a premium domain brokerage: a new animated mark, every page redesigned and rebuilt with designed mobile layouts, and a custom CMS for its sales record, featured names, press and enquiries.
Send us the structure of one technical file; we will show you the cross-checks we would run on it.
30-minute technical call · no deck