What the EU's move to data-driven conformity assessment means for your technical documentation

A new Horizon Europe project wants to replace PDFs and spreadsheets in MDR and IVDR conformity assessment with structured, machine-readable data. Here is what DIGICAP is, what it is not, and what you can sensibly do before 2029.

If you work in Regulatory Affairs or Quality Management, you have probably built a career on documents. Technical files, clinical evaluation reports, risk management files, all of them assembled with care, exported to PDF, and sent to a Notified Body who then asks for the same information again in a slightly different order. On 1 September 2026, the EU started funding a project to end that cycle. It is called DIGICAP, and it deserves a closer look than the average press release gets.

What DIGICAP actually is

DIGICAP stands for DIGItalisation of the Conformity Assessment Process. It is a Coordination and Support Action under Horizon Europe, funded with just under 4 million euros and running for 36 months, from September 2026 to August 2029. TEAM-NB, the European association of medical device Notified Bodies, coordinates it. Nineteen partners from thirteen countries take part: Notified Bodies such as BSI, TÜV SÜD, DNV and GMED, MedTech Europe as the industry voice, research organisations such as RISE and VITO, legal and public health expertise, and MDKU, the association that has been working on a unified data model since 2021.

Alongside the consortium, the project has invited a small group of regulatory software vendors to review and give feedback on the framework specifications. Five have been named so far: Veeva, DistillerSR, Aurevia, Hardian Health and meddevo. None of them sits in the consortium, none of them is funded by the project, and that is by design. DIGICAP describes itself as vendor-neutral. The data model is supposed to belong to the ecosystem, not to a tool, and the vendors are there to make sure it works in real software, not to own it. We take that role seriously, and we take the "not a member" part just as seriously.

The project's own diagnosis is refreshingly blunt. Certification under MDR and IVDR has become slow and hard to predict, and the reason is not the safety standard. It is the exchange layer. According to DIGICAP, certifying a device takes 13 to 18 months on average, and more than half of all applications arrive incomplete and trigger new review cycles. The information exists. It is just trapped in Word files, spreadsheets and PDFs that every party has to read, reformat and re-check by hand.

What it will deliver

DIGICAP has three concrete objectives. First, a consensus framework that describes a digital conformity assessment pathway, developed together with Notified Bodies, manufacturers and regtech software vendors. Second, a set of open-source, standardised data schemas and ontologies so that different tools can exchange technical documentation without losing meaning. Third, a scalable roadmap for rolling this out across Europe, one that the ecosystem has actually agreed to.

None of this stays on paper. At least four pilots will test the framework with real manufacturers and Notified Bodies across medical devices, in vitro diagnostics and software as a medical device. The project also runs a Stakeholder Panel of 30 to 35 external experts, including manufacturers and software vendors, that feeds into the framework through a structured Delphi process. All main results are to be published openly on the project website and on GitHub.

What it is not

This is where a bit of restraint is useful. DIGICAP is a coordination project, not a legislative act. It does not change MDR or IVDR, it does not create new obligations for manufacturers, and it does not force your Notified Body to accept structured data next year. It will produce a framework, schemas, pilot results and a roadmap. What the Commission, the MDCG and the Notified Bodies then do with those outputs is a separate question, and one that will take time.

So if anyone tells you that DIGICAP means you must rebuild your technical documentation by 2027, they are selling something. The same goes for anyone, vendors included, who claims to be building the framework. The honest version is that DIGICAP is the clearest signal so far about where the European system is heading, backed by the people who actually run it.

What RA and QM teams can do now

You do not need to wait for 2029 to benefit from the direction of travel. Three things are worth doing regardless of how DIGICAP turns out.

Start by asking where your technical documentation actually lives. Not the PDF you submitted, but the source data behind it. If the answer is "in the heads of two people and forty Word files," that is a risk today, independent of any EU project.

Then stop treating the exported document as the source of truth. Structured, reusable content that generates the document is the model DIGICAP is building toward. The earlier your internal processes work that way, the less painful any future transition will be.

Finally, follow the project. The framework and data schemas will appear on GitHub as they mature, and the Stakeholder Panel is open to manufacturers. If you want a say in how the data model for technical documentation looks, this is the moment to have it.

We keep a running summary of what has been published so far at meddevo.com/digicap

Our own contribution to DIGICAP is practical: what structured technical documentation looks like when RA and QM teams work with it every day, and what a manufacturer-side tool needs from the data model to be usable in a small company, not only in a large one. Separately, meddevo is a member of MDKU, the consortium partner leading the unified data model for technical documentation. We build meddevo eTD on the same premise the project is now formalising: technical documentation as structured data rather than as a stack of documents.