Health Tech Pathways and the Bottleneck Upstream of EU MDR
Health Tech Hub Copenhagen has launched an AI-guided platform to help startups navigate MDR. Useful.
But the more interesting lesson is what the tool cannot solve: founders who have not yet decided what product they are actually regulating.
Health Tech Pathways; HTH Copenhagen.
The most dangerous sentence in European health tech is not “we are building an AI medical device”.
It is: “We will deal with MDR later.”
That usually means after the pilot, after the pitch deck, after a friendly clinician has said something clinically useful, and after marketing has converted that into a claim with regulatory consequences. A classic startup manoeuvre: move fast, then discover that the paperwork was quietly moving faster.
On 16 June 2026, Health Tech Hub Copenhagen launched Health Tech Pathways, an AI-guided platform designed to walk health tech and medtech startups through EU MDR compliance: from the first question of whether the product is regulated at all, through to CE-marking preparation.
The platform is backed by the Novo Nordisk Foundation with DKK 32 million, around €4.3 million, and has been piloted over six months with more than 100 companies, experts and accelerators.
The launch story is tempting to write as “AI makes MDR faster”. That would be neat. It would also be too easy.
The better question is more operational:
Does Health Tech Pathways remove the MDR bottleneck, or does it simply move founders to the real bottleneck faster?
The MDR bottleneck is real
Europe’s MDR problem is not founder folklore told at accelerator demo days to frighten clinicians into hiring consultants.
The EU Medical Device Regulation has applied since 26 May 2021, replacing the previous medical device directives; the IVDR followed from 26 May 2022. The system is designed to raise standards for safety, transparency and post-market oversight. Worthy goals. Unfortunately, worthy goals do not make notified body calendars expand by magic.
MedTech Europe’s 2024 survey found that, for medical device manufacturers, average assessment times were 19.5 months for QMS assessment and 21.8 months for technical documentation assessment. More than half of conformity assessment time was spent in pre-review and certificate issuance phases, not the actual review of documentation.
The staffing picture is just as awkward. The same survey found that 91% of SME medical device manufacturers reported difficulty securing qualified regulatory affairs employees.
So yes, there is a real market need for tools that make MDR more legible. Especially for clinician-founders and digital health teams that know the patient problem but do not yet have a regulatory affairs function.
That is exactly the user Health Tech Pathways is targeting. The platform is explicitly built for founders who understand the product and the clinical need, but do not have regulatory expertise on the team.
What Health Tech Pathways actually offers?
Health Tech Pathways describes itself as an AI-guided regulatory platform for medical device software companies pursuing EU MDR or IVDR approval. Its public site positions the workflow around product description, MDR applicability, classification, documentation, consultant support and notified body discovery. The free trial includes qualification and classification, AI guidance, a consultant directory and a notified body directory; the Pro plan is listed at €299 per month excluding VAT.
That price point matters.
It sits between two bad options common in early-stage health tech: downloading a template and hoping for the best, or hiring expensive senior regulatory support before the product strategy is mature enough to use that advice well.
The more important feature, however, is not document generation. It is sequencing.
Most founders do not fail because they refuse to write documents. They fail because they write the wrong documents around an unstable product definition. Intended use, target users, risk claims, clinical benefit and evidence strategy are not admin details. They are commercialisation architecture.
And this is where the founder conversation becomes more interesting than the press release.
When asked where companies most often got stuck during the pilot, Jesper Grønbæk did not point to Annex IX, notified bodies or some obscure clause buried in MDR. He pointed to the product. Startups struggled to define intended use, clinical claims and target users. In his words, the bottleneck was often not regulation itself, but “product and strategy decisions that have regulatory consequences”.
That is the story.
BIO
Jesper Grønbæk
Founder & CEO, Health Tech Hub Copenhagen
Trained as an economist and management consultant, then helped software companies grow & scale for a decade, before mortgaging his house and being a software entrepreneur himself. Six years ago, Jesper devoted all his working time and effort to health tech.
Jesper Grønbæk,
Founder & CEO, HTH Copenhagen
The bottleneck is upstream of the bottleneck
You cannot classify a device whose intended use has not been decided.
You cannot scope a technical file around clinical claims that keep changing.
You cannot design an evidence plan if no one has agreed whether the product is supporting workflow, informing diagnosis, driving treatment, monitoring disease progression or merely making everyone feel digitally transformed.
This is why “MDR later” is such an expensive sentence. By the time “later” arrives, the regulatory pathway may already have been shaped by product, marketing and pilot decisions made months earlier by people who did not realise they were making regulatory decisions.
Health Tech Pathways is valuable if it makes those decisions visible earlier.
But operators should be precise about the value. The platform does not remove the hard thinking. It forces the hard thinking into the room earlier, where it should have been in the first place.
That is still useful. In fact, it may be the most useful thing about it.
Similar solutions exist, but most start later in the workflow
Health Tech Pathways is not entering an empty market. The regulatory automation and medtech compliance software landscape is already crowding, but the players solve different problems.
Flinn.ai is a visible adjacent European player. The Vienna-based company raised a $20 million Series A in February 2026 to expand its AI platform for regulatory and quality compliance across medtech and pharma workflows. Its focus is more on automating regulatory and quality workflows across the product lifecycle than on early founder triage.
Ketryx sits closer to regulated product development. It positions itself as an AI platform bridging development and regulatory compliance, generating traceability and documentation for medical device teams.
Rimsys and RegDesk belong more clearly in the regulatory information management category. Rimsys connects regulatory intelligence, product data, approvals, submissions, UDI and change management into a unified regulatory operations system. RegDesk positions itself as a RIM platform for medtech companies, combining regulatory data, submissions and live regulatory intelligence.
Greenlight Guru and OpenRegulatory’s Formwork address the QMS and technical documentation layer. Greenlight Guru covers quality, product development and clinical evidence workflows for medical device companies. OpenRegulatory’s Formwork offers a cloud-based eQMS for QMS and technical documentation, including risk management, usability tracking and documentation workflows.
Scarlet is not a comparable software tool, but it is strategically relevant. It positions itself as a notified body specialised in AI and Software as a Medical Device, which shows how the conformity assessment market itself is adapting to software-native products.
The distinction is important.
Most of these tools assume the company already knows it is regulated and broadly understands the pathway.
Health Tech Pathways enters earlier, at the triage gate:
Does MDR apply? What class are we likely dealing with? What work does that trigger?
That is a real white space for early-stage European digital health. It is also where mistakes are cheapest to fix.
What gets compressed is routine work, not accountability
The liability boundary is the part founders should read twice.
Health Tech Pathways’ terms say the platform is decision-support and educational tooling. It is not a notified body, competent authority, certification body, legal advice provider or medical device. Users remain responsible for regulatory decisions, submissions and outcomes. The terms also state that AI outputs may be incomplete, outdated or inaccurate and must be independently verified by a qualified regulatory professional before use.
Grønbæk’s answer to the same question was equally clear: the AI gives a “best estimate” on classification, but Health Tech Pathways does not take legal responsibility for that advice. The company developing the product remains responsible and should check classification with an experienced consultant where needed.
Good.
Not because disclaimers are inspiring — they are what prose becomes after lawyers remove the oxygen — but because this is the correct boundary.
AI can structure a first pass. It can explain requirements. It can reduce blank-page syndrome. It can make founders ask better questions before they burn six months building in the wrong direction.
But it cannot own the intended use.
It cannot own the clinical claim.
It cannot own the risk-benefit argument.
It cannot stand in front of a notified body and defend the evidence package.
That accountability remains with the manufacturer.
The consultant market changes, but does not disappear
The lazy take is that AI regulatory platforms replace junior consultants.
The more likely outcome is re-segmentation.
Routine triage, document structuring, requirement mapping and first-pass education become cheaper and more self-serve. That will hurt consultants whose business model is charging senior rates for basic orientation. Some business models deserve a brisk walk off the pier.
But the valuable consultant becomes more important, not less.
The work shifts towards strategic review: validating classification, challenging claims, stress-testing evidence, advising on notified body selection, and spotting the grey-zone risks that a founder will miss because the product still looks obvious to the person who invented it.
Grønbæk’s own framing is that the platform should make startups more self-sufficient on early-stage tasks, while consultants focus on complex areas and regulatory grey zones.
For founders, that changes unit economics. Not because regulatory expertise becomes cheap, but because expensive expertise is used where it matters.
For consultants, it changes positioning. The first-pass work gets commoditised. The strategic judgement becomes the premium layer.
The AI Act makes the operating model more important
For AI-enabled health tech, MDR is no longer the only regulatory track to watch.
The EU AI Act entered into force on 1 August 2024. The European Commission says high-risk AI systems, including AI-based software intended for medical purposes, must meet requirements around risk mitigation, high-quality datasets, clear user information and human oversight.
This means founders are not moving into a simpler world. They are moving into a more layered one: MDR or IVDR, AI Act, GDPR, cybersecurity, clinical evidence, QMS, post-market surveillance and procurement reality. Delightful. Like a wedding cake, if every layer required legal review.
That makes the case for better early-stage regulatory operating systems stronger. It also makes the “AI assistant plus human expert” model more credible than the fantasy of fully automated compliance.
The future is not AI replacing regulatory accountability. It is AI making accountability harder to ignore.
The quiet asset: the map
There is one strategic question Health Tech Pathways did not answer directly.
When asked whether the Novo Nordisk Foundation-backed platform could become a data layer for European regulatory harmonisation, Grønbæk said he was not in a position to comment.
Fair enough.
But from an operator’s lens, the possibility is obvious.
A non-profit, ecosystem-backed platform sitting at the classification gate could accumulate a structured picture of where startups get stuck, which products fall into grey zones, what intended-use patterns create classification problems, and where human consultants are most often needed.
That does not mean Health Tech Hub Copenhagen is building a policy influence engine. It does mean the most valuable asset may not be the workflow interface. It may be the map of how early-stage European health tech actually collides with MDR before anyone reaches a notified body.
That map would be useful to founders, accelerators, consultants, investors, regulators, and policy teams.
In Europe, that is what strategic infrastructure often looks like: not loud, not flashy, and quietly more important than the pitch deck suggests.
What operators should watch?
When startups hire regulatory leadership?
Consultant directory
Whether accelerators make this a standard operating layer?
Product-definition gap
One Thing to Remember
