João Pedro Ribeiro, CEO & Co-Founder @ PeekMed, shared with us the navigation challenges of regulation in healthcare and AI, his experience in expanding the company to different markets and their focus on innovation. This is a story of persistence, adaptability and the art of balancing innovation with compliance.
PeekMed is the first fully automated AI-driven system that simulates orthopedic procedures and drives efficient execution, which aims to decrease the cost of healthcare, while helping surgeons improving outcomes.
The genesis
João Pedro Ribeiro, a biomedical engineer with a fascination for medical imaging, teamed up with co-founders Sara Silva and Jaime Campos to commercialize AI-driven 3D surgical planning software born from Ribeiro’s graduate research. They wanted to help surgeons to plan procedures with data-driven precision, instead of the “eyeballing” approach they used, and now that’s effectively what they do all around the world.
Orthopedic surgeons faced mounting pressure to reduce costs and improve outcomes, yet tools for preoperative planning lagged behind other specialties. PeekMed decided to focus on helping orthopedic surgeons to plan their surgeries. By 2023, PeekMed became one of the first Portuguese AI health companies to secure FDA clearance, a milestone that opened doors to the U.S. market.
In the beginning, the main challenge for PeekMed was access to data. It was necessary to create partnerships with hospitals (CUF, for example) to get access to the data that would allow them to train the model. João says “it was the simplest to overcome, although simplicity is debatable”.
The next challenge was related to the business model itself. PeekMed was changing a paradigm in the orthopedics market. Doctors and hospitals were very used to pay for a license of the products, many times a lifetime one, and they would only pay extra if needed maintenance or an extra feature. For context to the readers, PeekMed exists for more than 10 years. They wanted to introduce the SaaS (Software as a Service) model, in which you pay while you use, almost like a lease of the product. This was also a challenge in their particular market because it wasn't common. It was difficult to explain to hospitals how it was an advantage, an added value and why it didn't necessarily mean that the product would become more expensive.
Naturally, the regulatory topic was also a challenge, in the European market and the North American market - they're in both.
Regulation struggles
PeekMed’s rise coincided with a seismic shift in medical device regulation. In 2021, the EU’s Medical Device Regulation (MDR) replaced the outdated MDD, imposing stricter requirements for AI-driven software. For Ribeiro, this regulatory overhaul exposed a critical flaw: Europe’s “stop-and-stare” approach to innovation.
“In the U.S., regulators work with companies to clarify expectations. In Europe, when they don’t know the answer, everything grinds to a halt”
- João Ribeiro explains in a conversation with us.
The regulatory challenges intensified as the demand for MDR certifications exceeded the number of EU-approved bodies that could provide them. Nowadays almost every software product, specially the ones with AI, are consider medical devices, which mean they need certification, and startups launching AI health tools in Europe face wait times of 9–12 months and costly certification — resulting in a “bottleneck to innovation” says João. In addition, when the regulation changed, certified organizations lost their certifications, making the problem even worse.
The rules must exist. Regulation needs to exist. In the end of the day, “I’m a patient and I want the guarantee that everything that’s used to treat me was verified by a lot of people”, he explains.
But it shouldn’t limit how companies grow.
From his vast experience operating a health care product in Europe, he understands that main problem lies within how to react to uncertainty, and not the regulation itself. It’s not yet clear, for the regulatory body, what information they need to certify the product as being safe and accurate. There’s a vast list of topics they can cover, from data access to models’ accuracy, but the process is still unclear. The way Europe seems to react to doubts is to stop the process. As a company you don’t have visibility to what’s going to happen next and so you’re stuck. Innovation stops.
In the United States of America, there seems to be more clarity of what they are looking for. If there isn’t, it’s an evolutionary process between the company and the agency.
The Food and Drugs Administration (FDA) works as the certifying and regulatory body, which makes the process must faster. In Europe, the current entities for certification and regulation are separated, so even though a product is certified for the European Union and for all state members, it still needs to be register in every local entity in each country that sells the product in order to be regulated by it. Whereas in the United States of America, a company would only do it once with the FDA, who certifies and regulates, and it’s done for the 50 states. This is such a big topic that the European Union is discussing changing their model to a similar of the USA. Although the FDA process itself has its inefficiencies, João states “I'd prefer the FDA process to the European model”.
International Expansion
PeekMed started in the European market outside Portugal but then went to the United States. When expanding to different markets it’s important to adapt the business model accordingly.
“The way we sell, for example in the European market, is different from the way we sell in the North American market.”
- João mentioned.
Their main European market is the German market, in which they sell directly to hospitals and doctors (mainly private clinics). This was a market that was very used to licenses and the SaaS logic was a challenge. Whereas in the United States they sell through manufacturers to distribute it to their clients and the SaaS model was not even an issue. The business model itself is different (end customer or distribution partners) although the way of accessing the product is the same.
“So having clarified the regulatory issue in the different markets, it has much more to do with how we sell the product, how we distribute it and the business model than the product itself. Understanding what the market absorbs most.”
The future of healthcare
Having been asked about the future of healthcare before, João Pedro Ribeiro shared with us his vision of a more preventive and proactive medicine, rather than the reactive approach we’re so used to. “Typically, you don't go to the doctor to see that you don't have a problem”, he explains. Because it’s the way we see healthcare, artificial intelligence also works in a reactive way - “you send it an image and it analyzes it”.
Imagine algorithms predicting a patient’s osteoarthritis risk years before symptoms appear—and recommending lifestyle changes to avoid surgery. For that we would probably need models who are running and who evaluate your body 24/7 and who provide suggestions based on that collected information, and even maybe genetic.
Preventive medicine is also much cheaper for a state since individuals will be able to prevent health conditions (they won’t require treatment) and it’s going to be much cheaper.
A clear problem is regarding access to data. Even though, as an individual, I can have gadgets that collect the data and provide me insights, there will be a time where a visit to the doctor is necessary. How can someone share the information collected with their doctor? How can a doctor or clinician follow my history? “For it to work, the doctor, the clinician, has to be part of the process”, João states, “We're still going to have to develop and demystify this access to data where we're very attached to it”.
What excites more João Ribeiro about PeekMed it’s very much related not only to the systematization of processes, but also to achieving reproducible, efficient and effective processes.
Effective, meaning without the need for new surgeries, i.e. surgery done right the first time.
Efficient - managing to do a good surgery in the shortest possible time at the lowest possible cost.
Reproducible - having different patients with different treatments, and still managing to reproduce good results across all surgeries. It’s answering the question: “How do I guarantee that a doctor who has 8 surgeries a day, with 8 different preparations because there are 8 different patients, but the practical results are all excellent?”.
Final Thought
PeekMed’s story underscores a truth too often ignored in the AI hype cycle: Technology alone doesn’t transform healthcare. It takes resilient entrepreneurs willing to decode regulatory labyrinths, rethink business models, and—above all—listen to the humans their tools serve.