In this episode, we will answer the following questions:

[01:26] a large platform for the new generation of eye disease treatment

[09:50] how ML helps in the diagnosis and treatment

[12:56] ML accuracy and bias in making a diagnosis

[14:35] the process of hiring optical engineers

[18:57] in-house vs. outsourced development

[19:42] product distribution plans

[21:23] go-to-market company strategy

[24:05] how important are soft skills for engineers and how to evaluate them

[26:04] Geoffroy’s scientific works in orthopedic

[29:44] sales process: making the right message for the right person

[33:23] Geoffroy’s advice on becoming a good skilled engineering manager

[38:33] Rapid Fire Round (3 personal questions)

 

Links:

Geoffroy’s Linkedin: https://www.linkedin.com/in/geoffroy-rivet-sabourin-b1007920/

PulseMedica website: https://www.pulsemedica.com/

 

Listen to the episode on iTunes, SoundCloud, Spotify, YouTube, Google Podcasts and let us know what you think on the topic.

The mission of our podcast is to show the real-life challenges of implementing technology in healthcare and the podcast is sponsored by Demigos, a company that develops custom IT solutions for healthcare startups and companies.

My name is Ivan Dunskiy, your host is always, and I am joined by the guest with an interesting background, Geoffroy Rivet-Sabourin. Geoffroy is the CTO at PulseMedica. It's a medical device platform that enables effective and noninvasive treatment of various eye diseases.

Geoffroy trained as a computer engineer with the Ph.D. in medical image processing and in the last 15 years has developed deep expertise in medical device software development. And before PulseMedica he helped multiple companies to bring their medical devices onto the market. Geoffroy, thank you for joining. How are you today? 

Geoffroy Rivet-Sabourin: I'm doing good pleasure to be with you today. 

Ivan Dunskiy: Okay, let's start with a brief explanation of what PulseMedica does and like what products you create.

[01:26] a large platform for the new generation of eye disease treatment

Geoffroy Rivet-Sabourin: We are developing a large platform for the new generation of eye disease treatment. I like to say that currently a lot of treatments for eye disease based on lasers are really surgeon-dependent because the surgeon is driving the laser mostly manually with some apps. But the technology we are developing is a computer drive laser, and mostly partly automate or automate the targeting of some specific locations in the eye to treat them with the laser at the end with more accuracy, more repeatability for the doctor and also increased the volume of a patient a doctor can treat in a day. 

Ivan Dunskiy: Okay. And the technology already works in the market, right? 

Geoffroy Rivet-Sabourin: No. We are currently in the development phase. We have just recently received authorization from Health Canada because we are based in Canada. For the first trial with one of our devices, we expect that we will be in a process of developing the technology for probably the next 24 months, at least. But on the way to this road, we will probably have one or two products, but not necessarily for treatment but more for diagnosis purposes because we have a huge machine learning team also working on the planning process. But for the first treatment device, we expect probably to go on the market on the first trial in 18 - 24 months. And for the product a little bit later. 

Ivan Dunskiy: And I'm not really familiar with the medical device production cycle. My experience is more of the software development side. So could you please educate us on what was the approach of developing the device and how you enabled machine learning? At what stage did you start the software development? Just elaborate on the whole product life cycle.

Geoffroy Rivet-Sabourin: Yeah. This is important to understand that the PulseMedica product is a large variety of technology. We are developing the device completely from scratch, meaning that we have some optical engineers on board to develop all the optical support. We have electronics, firmware, and embedded software for the machine learning team. 

The life cycle of the product is basically trying to move most as in parallel as possible before all the components of the device, but it sure that on some aspects, some software pieces can go to the market faster. We need to think about that all the hardware is probably a little bit trickier or more difficult in terms of approbation, just thinking about all the electrical safety, you know, you need to comply with a lot of electrical standards and rely on external vendors to do some audit on your product. And this kind of process is taking long and can be done only on the final, final product, because it's costly. It takes time and it is costly, meaning that you will not do that in multiple times during your development, you will do it only when the final version will be ready. This is why this aspect of hardware can take a little bit longer than, for example, only software cause just the software itself, you compile your software, deploy your software, do your verification and validation and you are mostly ready to go. 

Ivan Dunskiy: And do you go through the clinical trial period? 

Geoffroy Rivet-Sabourin: Yeah, we are now starting a clinical trial for our first device on a limited number of patients. This is the first trial for 60 patients in Canada. We are already envisioning, another larger trial, probably us based in the next 12 to 18 months. In this case, we are looking, the simple size is not totally defined for now. It's probably a question of 500 to 2000 patients distributed between 5 to 10 sites in the United States. And in probably around 24 months we would like to do a first pilot study on our treatment device. That will be in this case a way a limited number of patients just to demonstrate a proof of concept of the technology in their real life.

Ivan Dunskiy: And would that be enough to after that period that you will have some first version of the product that you will be ready to launch?

Geoffroy Rivet-Sabourin: For some aspects of diagnosis devices we can probably be ready to go to the market in this case, but for the treatment device it will not be enough data to say that we can convince or sell and also FDA or any other regulator that the device is safe and efficient to go to the market.

Ivan Dunskiy: And for the treatment device how long do you anticipate that this verification period will last?

Geoffroy Rivet-Sabourin: Depending on the result we will get. But this is a question of probably a year and a half to do that. And the risk. We need to think about that. The risk for the patient is way higher. It is always depending of the patient's risk. The more the risk of the patient is large, the bigger is the verification evaluation plan. 

Ivan Dunskiy: Yeah, understood. And what part of the whole product the software plays? Is it something that plays a supportive role in the whole product, or is it much bigger than that?

Geoffroy Rivet-Sabourin: I will say that it is a large piece of PulseMedica technology. A large piece of what we are doing is software, the know-how of the company. And I will say the big quality of what we are developing is really related to the software we are currently developing.

Ivan Dunskiy: So is your focus more on the device part or on the software part?

Geoffroy Rivet-Sabourin: Both at the same time, because it is sure that we need some cutting-edge technology to take images for the kind of technology we are developing. But at the same time, we need to extract all the information in real-time from this data which require a lot of development in term of image processing and machine learning.

Ivan Dunskiy: And so you mentioned the eye diseases. Could you please elaborate on what categories of diseases the device will help with? 

Geoffroy Rivet-Sabourin: Yeah, it is a question of at first application. What we are looking for is something that will be able to be applicable for a lot of diseases going to dry MD, glaucoma and retina surgery. The first application and where we would like to do a proof of concept is really related to photocoagulation, for diabetic retinopathy. That means doing some kind of photocoagulation on the retina. This is the first application we are targeting. Why this application, because it fits well with the kind of images we would like to do. We have also already developed a lot of pre-planning things around that with machine learning. And we see that it can be a good treatment for the patient to avoid monthly injections and have a better life for patients.

[09:50] how ML helps in the diagnosis and treatment

The idea is to help the doctor because ML can do two things. We have one aspect of machine learning that is taking care of speeding up the doctor process. If the doctor has to select 100, 200 points on a patient retina, it can be a really long job and take a long time for the doctor and cost money at the end to do that. This is one aspect, speeding up the doctor process. 

But the other thing we would like to do is also to be able to predict the evolution of the procedure, the patient disease, to help the doctor to select the right treatment for this patient or selecting the right treatment that can be at some point with no treatment. If the doctor realize that based on the data that the disease of the patient will not evolve so much, and this is not necessary to treat. You need to keep in mind that treating with a photocoagulation laser, the retina of a patient is partly destroying the retina on some specific location. And if you don't have to treat, this is probably better for the patient, but a lot of those patients need treatment to slow down the evolution of the disease. 

Ivan Dunskiy: Yeah. So basically it will help to decide to be more sure whether a patient needs treatment or not?

Geoffroy Rivet-Sabourin: Exactly on some. And retinopathy is one of the treatments, but we are targeting some other pathologies. This is difficult now for the doctor to make a decision of intervention or no intervention. And we would like to help them, the idea of machine learning is to give a longer perspective on the data of the patient and the history of the patient.

Ivan Dunskiy: So does it mean that in your machine learning algorithm you look through big amounts of data for a specific patient rather than comparing into to other patients with similar cases, right?

Geoffroy Rivet-Sabourin: We do both. We are able to look at the data on a single patient from the historical imaging of the patient. But we are also able to include. And what about a doctor treating those patients with similar pathology in other cases, what was the result and for a large database. 

Geoffroy Rivet-Sabourin: In this case, you understand that this is not just analyzing images. This is more than analyzing images. This is also going in doctor notes or clinical data, more than inside of EMR, for example.

[12:56] ML accuracy and bias in making a diagnosis 

Ivan Dunskiy: It’s interesting. Is it a way to evaluate how machine learning suggestion is accurate? Because I think that on the negative side of things, it can bias the doctor's decision. So, what do you think about that?

Geoffroy Rivet-Sabourin: It is exactly a big challenge and only experience and evaluation across clinical data or clinical trial can really give good advice about that. It is sure that we can do some retrospective evaluations saying that: “okay, I'm blindly using this patient that I already know the treatment, but I'm not using the recent treatment. And I'm stopping at time. For example, my analyzing. Times zero, but I'm just looking at minus 1, 2, 3 for training or evaluating this patient and look at, okay.” In reality, this patient has already be treated. And the systems say exactly the truth according to this, the real life of this patient, but this is really difficult to do that. And this is why some regulators like FDA are still questioning this kind of thing and know that these kinds of things present some challenges. 

Ivan Dunskiy: Yeah. Because you know, there is kind of maybe on the moral side of things or on the decision-making side of things. So when a doctor makes a decision, that is his responsibility, right? But now with the machine learning algorithm, who is who, who is responsible for that? Obviously, maybe there is a doctor, but.

Geoffroy Rivet-Sabourin: Yeah, but the device is helping drive the doctor’s decision. You are totally right. This is a big challenge.

[14:35] the process of hiring optical engineers 

Ivan Dunskiy: Yeah. And you mentioned that you have optical engineers in your team. I assume that is quite challenging to hire those specialists. The labor market of engineers is quite hard and I assume that as for the optical engineers, it is much more difficult than just the average engineer. So what is your approach? What are your thoughts on hiring your team and basically how you do that?

Geoffroy Rivet-Sabourin: You are totally right. Hiring engineers is currently a big challenge. Probably not just for us, but for everyone. Mostly now we are currently looking for software engineers. This is challenging, optical engineers are also challenging. How we overcome that? Basically we are a really open company. We are open to recruit mostly around the world to hire the best engineer as possible. But we have chance to have also locally in term of in the west of Canada, having a really great institution to train engineers, mostly optical engineer. We are based in Edmonton, Alberta where we have an excellent university there in optic. That's really helpful. There are some other great universities in Ontario training this kind of people there. Basically, this is a mix between local resources. We have hired employees from Japan directly who came to Canada to help us. This is probably the secret to being open, open minded and be ready to hire the best people from everywhere. 

Ivan Dunskiy: And do you work remotely? 

Geoffroy Rivet-Sabourin: Yeah, we ideally depending on the position we are open or not to, or open less, I will say, we're not completely close to remote work for sure. Every employee can do some remote work, but we have some employees that are mostly remote. It fits with what I just said. If you would like to recruit everywhere, you need to be ready to do that. It's bringing some other challenges remote work, but we're realizing the last two years that it can be feasible.

Ivan Dunskiy: What is your approach of integrating medical device team, with ML team, with software development team? Is it the same like you integrate different software development teams together or it is completely different? 

Geoffroy Rivet-Sabourin: No, it is similar. This is just a question of synchronizing teams together and be sure that everyone know, understand needs of the other team is mostly real when you would like to synchronize electronic team and software team. This is not so big challenge. This is just to be sure that you have good communication with your team and between your teams and everyone knows what they have to do.

Ivan Dunskiy: So then it is about hiring right people and having a good culture in the company. 

Geoffroy Rivet-Sabourin: Yeah, and be sure that people understand requirements of every discipline, if software know the limits of electronic and electronics know the limit of software. And in this way you can have good complimentary work between teams. 

Ivan Dunskiy: Yeah. So it comes to experience. And how big is your team? 

Geoffroy Rivet-Sabourin: We are currently 25, we have a large project starting with what I’ve just discussed in the next weeks. And we are currently working really hard to increase the size of the team.

[18:57] in-house vs. outsourced development 

Ivan Dunskiy: And will you hire only in-house or are you looking for hire from third-party vendors, and software development partners?

Geoffroy Rivet-Sabourin: We currently have a few subcontractors. Honestly, we clearly prefer to have internal resources for multiple reasons. We are open to third-party vendors in the context of specific modules, for example, but to develop the core code of the business, we will certainly rely more on internal resources. Just because this is mainly the know of the company. 

Ivan Dunskiy: And I assume that is your thought process why you want to have internal teams just to keep the kind of know-how and knowledge inside the team, right?

Geoffroy Rivet-Sabourin: Exactly. This is exactly right. And this is important for us. A lot of the IP of the company that is valuable for the company is in the ends of those people. 

[19:42] product distribution plans

Ivan Dunskiy: And what are your plans for the product distribution? Do you plan to distribute it in Canada or in the US as well? 

Geoffroy Rivet-Sabourin: As a lot of companies or natural way to go to US, this is basically a question of market size and attraction. This is probably the first goal. Instead of in Canada where it is sure that we will do some things in Canada and try to, and have our product cleared in Canada. But this is just a question that the US market is 10 times the Canadian market. And you need to think about that. It is mostly the same regulatory reward to be approved in Canada and the US. We’re just talking about Canada and the US, but after that, you can think about Europe, but it is another part of the word to be approved in Europe. This is a large market, but we need to start somewhere. We will start probably by the most interesting one, based on the market site compared to work need to enter this market. This is probably the way to go.

[21:23] go-to-market company strategy

Ivan Dunskiy: Can you share, what is your go-to-market strategy? Will you go to ophthalmology practices or to large hospitals? So what is the plan? If you can share. 

Geoffroy Rivet-Sabourin: Yeah, this is honestly we are not at this point for now, but a big hospital is probably not the first market for us. Probably doctor clinics are probably the first place to go for us.

Ivan Dunskiy: Who provides specifically that type of service. 

Geoffroy Rivet-Sabourin: Exactly. Hospitals sometimes are more challenging. This is a kind of device that is fairly easy to sell directly to clinics. And also it's clearly related to the practice of ophthalmology which is something that is really practiced directly inside of the doctor's clinic and doctor's office and this is important to understand too, that the device we are creating is more office based. 

Ivan Dunskiy: Great. And do you already plan with the scaling? So let's say that at some point you will face a lot of demand. So do you already have a plan how you would meet that and would work with that in terms of producing devices that is much easier, I think, on the software development, on the software side, but yeah. What is your plan?

Geoffroy Rivet-Sabourin: Yeah, this challenge is not super difficult to address in terms of software because it is relatively easy to multiply software. It is more challenging when you have hardware. In the short term, we're not planning to manufacture the device ourselves, probably partner with a contract manufacturer or this kind of group can be certainly beneficial for us. Putting together a manufacturing plan for this kind of device is certainly really challenging and costly and probably not our core business. This is why for now, I'm not saying that it will not change. But for now it's probably partnering with the group for the hardware part.

Ivan Dunskiy: As for the software, I assume that you have a platform for doctors. Do you have any products for patients as well?

Geoffroy Rivet-Sabourin: Not for now. And it is really something that will be more of an indication for doctors. We are thinking about the long-term future, some kind of home based device for diagnosis proposed, but I cannot say this is not, this is just ideas, brainstorming ideas, not necessarily something that will be that we have really good definition.

[24:05] how important are soft skills for engineers and how to evaluate them

Ivan Dunskiy: And having 25 engineers in your team, how important do you think are the soft skills to make the work productive and what is your way to maybe evaluate those skills when you recruit people and increase the level of that skill during the work?

Geoffroy Rivet-Sabourin: Yeah, this is a really good point. Evaluating skills. When we are in the hiring process, it’s challenging. I'm doing multiple interviews per week at this time. And hiring a person is also a tough process. Because you would like to be sure to make the right choice. Sometimes you have few candidates, you would like to pick the right one. We are focusing on skills for sure. This is important because we have technical work to do. We are trying to have some kind of technical interview trying to put the person most fast as possible in the context of the real work. But we are also thinking that person’s attitude and behavior, the person is important to have the right fits to continue. The other part of your question is how we would like to improve those skills. As having a good fit of the person in the team will make this person benefit as most as possible of the team work and learn from other members of the team. We are also trying to have different levels, junior to senior people, and we are expecting to increase the level of our junior people based on the experience of seniors. And at the end, we are also completely open to have people to go to any training that can be relevant for their work and improve the quality of our people.

[26:04] Geoffroy’s scientific works in orthopedic

 Ivan Dunskiy: Cool. And I also noticed some of your works on the Research Gate portal and also that you have a patent for TBL processes. Are you still involved in the scientific work? Or you mostly focused on PulseMedica?

Geoffroy Rivet-Sabourin: Yeah, I have multiple patents related to orthopedics because I'm coming from this world before PulseMedica, I've been in orthopedic companies for 10 years. In the context of this work in orthopedic, I've worked with multiple universities for developing some cutting edge technologies. At PulseMedica, we don't have precise current collaboration or formal collaboration with university, but we have multiple of our members that are mostly for machine learning that are really tied together with some super cutting edge group in machine learning, mostly in the University of Alberta.

Ivan Dunskiy: Does your previous experience help in your current work?

Geoffroy Rivet-Sabourin: Yeah, for sure you, it is always a piling experience. What is important is this orthopedic company was also part of this software medical device. This is where I've created a lot of my expertise learning also how to bring some product regulators is certainly an understanding of the development process is probably what's is my expertise now. 

Ivan Dunskiy: So to understand the whole process, not just how to develop something, but to bring it to the market.

Geoffroy Rivet-Sabourin: Yeah. Under understanding the process is understanding the process from understanding the market and what the medical device market would like to have. You will say that what is the relationship between orthopedic market and ophthalmology margin market? Totally. This is totally two different market, but what you learn from one market, you know what you need to look at to understand this market and understanding the processes, understanding the market, basically user needs.

And after that, how to develop this kind of product and mostly how you can go to the market. By the way of regulator, how to approach this process of VNV and preparing documentation, preparing the process and also to have a good strategy for the regulator, because you can develop the most beautiful product but if you are not able to go across FDA or any regulator, because it is too innovative or it is too different, or you have a lack of data, you will not be able to touch the market with it.

Ivan Dunskiy: So at the end, you need to understand the regulation process to bring kind of the right product to the market, right?

Geoffroy Rivet-Sabourin: Exactly. Exactly. Because some companies failed just because they are trying to bring, I will say, this is bad to say that, but they are trying to bring too much innovation at the time. The medical device is a question of baby steps, small innovation piling, some few small innovation. One after the other is you will touch and go to the market probably quicker than trying to do a big hit with disrupting innovation. 

[29:44] sales process: making the right message for the right person

Ivan Dunskiy: Yeah. My assumption is that the thought process of a regulator is to mitigate risk. That is kind of their focus and as an entrepreneur, as an innovator, your thought process and your aim is to innovate something new that would completely disrupt the current, how things will work. So, and here you need to find that balance between those two things.

Geoffroy Rivet-Sabourin: Yeah, you're totally right. This is exactly that. And what I'm saying all the time to people is there are two for the same product. You have two sale pitch, one sale pitch for the marketing people that you are the most beautiful, innovative product ever. And for the same exact product for regulator, the sale pitch will be “we are the most boring product ever, we are doing exactly the same thing as others.” But this is exactly the same product. This is just making the right message for the right person. But this is why you cannot be the super cutting edge technology, super innovating technology and convince the FDA that you are boring product. The question is you just need to have enough innovation to make a difference on the market and make a difference for patients.

Ivan Dunskiy: Yeah, we are the same like them, but we are more boring and more safe. 

Geoffroy Rivet-Sabourin: Exactly, exactly that. But you are totally right. FDA is caring about patient risk and efficiency. You need to put on their shoes and understand their way of thinking. 

Ivan Dunskiy: And as a startup, that maybe didn't have experience dealing with reg regulators, what do you think is the method for how they can understand what would work for regulators? 

Geoffroy Rivet-Sabourin: This is not so easy but this is because it can change from product to product. That's clear, it's really to analyze the data. You have to demonstrate that our strong is the demonstration of safety and be sure to have good rationals about what you are putting in front of FDA. But this is difficult to make a rule around that every product will be different. And it is also depends on how different you are from what is currently in the market. And possibly sometimes one strategy is to say, okay, we have this big project, but how we can cut it in pieces to have just small improvements and demonstrate some small improvement and go multiple time to regulators by demonstrating small improvements.

Ivan Dunskiy: So to show real-life evidence of that the technology works.

Geoffroy Rivet-Sabourin: Yeah, and this real-life evidence is important to be exposed to the market sooner than later. For a product, this is really important. Because as an engineer, we can develop super nice products but forgetting a lot of things that the real market will tell you, we can work. And we are always working with doctors as advisors, but this is a small sample set of doctors. 

[33:23] Geoffroy’s advice on becoming a good skilled engineering manager

Ivan Dunskiy: That is maybe the last question I would like to ask from this set of questions. So it is clear that every person who was a manager or was involved in scientific work, and then became engineer manager, each person has his own path. So what kind of advice can you give to those who want to become good skilled engineering manager?

Geoffroy Rivet-Sabourin: I think the first one and probably the most important one is don't be scared to have the best people around you and having the best people around you will right away bring you to delegate some stuff and delegate some knowledge and be sure that you don't know exactly everything but you have the best with you to know that. For example, if I have a super skilled machine learning engineer, I'm not the super skilled guy in machine learning, but I have with me in my team, the best one. And this is how we can have a good team and have the same thing for electronics, the same thing for software. I have the best one, but you, as a manager, can be humble and say: “Okay, I don't know everything, but you have the best with me, and they know.”

Ivan Dunskiy: Yeah. So basically to improve your hiring and recruitment skills, as well as your ability to delegate and trust to people you work with.

Geoffroy Rivet-Sabourin: Exactly, exactly that. And after that, when you have the best, always the best you can have, these guys will also hire the best for their junior engineers, because these guys will not be scared to be not the best guy. And this is again, why attitude is really important. 

Ivan Dunskiy: And that creates this kind of culture that grows. So once you have your example of how you hired people, then people who hire other people, they kind of cultivate the same culture.

Geoffroy Rivet-Sabourin: Yeah. And I think people that put the team before them in terms of priority for the company and be sure that they will understand the success of the team is more important than their own success, because the success of a single person, mostly when you have multiple technology project, you can be super successful and having super nice electronic for your project. But if you don't have the optic, you don't have the software, you have nothing. And it largely after that, if you go outside of development, if you have a super nice project and super nice product, but we don't have super nice people to sell it and put it on the market, you know, a good strategy to go there. You have nothing in turn of business. This is teamwork, and this is why having people, and putting the team in front of everything is the most important thing. 

Ivan Dunskiy: Yeah, there is also a question about how to show that bigger picture to specialists. And I think there is a responsibility of a manager because engineers are often focused on their craft, on writing code or making models or making medical devices. And they are good at that and that is kind of their passion but then when they work in a team, they need to see a bigger picture, and seeing a bigger picture, they understand the value of other team members. And that's how they come to that thought as you mentioned that the team is in front and like their priority. 

Geoffroy Rivet-Sabourin: Exactly. And at PulseMedica we have really great chance to have founder/CEO is a nice person, is one of the best I've seen to be able to put this culture inside of the company bringing a nice team together. And this is right for engineer. We have also fantastic people on a doctor advisory board, fantastic group of people on a board of director. And this is mostly who put all these people together and make it the best. And also is someone that is helping a lot to put a lot of transparencies and in the company and having people knowing the right thing about the company helped them to be really involved inside of the project.

[38:33] Rapid Fire Round (3 personal questions)

Ivan Dunskiy: Yeah. Thank you. I think that we covered a lot. So I would like to ask you several short questions on the personal level. We call this exercise Rapid fire round. So what is the latest movie that impressed you the most?

Geoffroy Rivet-Sabourin: Honestly, this is not a big movie, but the last “Top gun” was interesting. This is probably a question I've saw the first one when I was younger and it's probably a return to the child part of my life.

Ivan Dunskiy: Okay. I haven't seen it yet. And what’s your favorite book?

Geoffroy Rivet-Sabourin: Recently I have limited time for reading, but I'm mostly reading some technical books. One of the latest, latest one is about software embedded validation. That was interesting because this book is giving really nice key for people or company who would like to start this process.

Ivan Dunskiy: And could you elaborate on this? Like, what is specifically about?

Geoffroy Rivet-Sabourin: Yeah. If this book is giving some good advice about when you are a company, you have a medical device software, and you would like to start your validation process. Where do you need to start when you have no experience. And this book is really interesting one, giving good initial thought about what is the process and what is the guideline you need to follow. 

Ivan Dunskiy: Got it. And what is one piece of advice you would give to your 20 year old self? 

Geoffroy Rivet-Sabourin: The one I preferred is you are young. You have no, some engagement in life. Be part of a startup because this is a nice place to learn. Sure you need to have the good mindset, but you will learn so much inside of a startup, have access to various responsibilities. If you are willing to try new stuff, startups always have new challenges. You just have to raise your hand and go ahead and take the challenge. And I did that in the past. I was basically in the orthopedic company, I was hired as a software engineer to develop the core engine for image processing. But after 18 months in this company, I was starting doing design of clinical trial protocols, do mechanical designs and lead the mechanical engineering team. This is something that is more difficult if you are hired inside of a big organization, but you need to be comfortable to be uncomfortable.

Ivan Dunskiy: Yeah, for sure. That's precisely said. I think that is a perfect way to end today's interview. So, thank you for sharing a lot about the way how you explain how different teams work together was valuable as well as the process how important it is to understand the mindset of regulator, if you want to bring a product that requires that on the market. It is really beneficial for those who are starting their journey. Before we finish, what is the best way to get in touch with you if people want to connect? 

Geoffroy Rivet-Sabourin: By the way, the website of the PulseMedica is one way, but also people can try to connect with me on Linkedin. I'm always pleased to make new connections there and meet new people there.

Ivan Dunskiy: Thank you, listeners. And we'll catch up on the next episodes.

Who is behind the HealthTech Beat podcast

We are a team of IT professionals who like sharing technical knowledge with healthcare industry people.

At Demigos, we generate ideas on how to improve product performance, design, and positioning based on our experience building complex health tech solutions.

Check our blog with articles on the related topics, and our cases in healthtech. Also, connect the podcast host and the CEO of Demigos Ivan Dunskiy on Linkedin.