S5E17 AI Meets Dermatology: Jonathan Benassaya

A near miss changed everything. When a dermatologist almost sent Jonathan Benasaya home during COVID without checking under his mask, a hidden melanoma forced a reckoning with how we detect skin cancer: slow access, manual exams, and invasive treatments that arrive too late. That moment now fuels SkinBit, a clinical‑grade, full‑body scanner built to triage quickly, track change over time, and help dermatologists focus on the patients who need them most.

We sit down with Jonathan to unpack the big idea: use dermatoscopic‑resolution imaging to create a digital twin of your skin, then apply AI to score suspicious lesions and prioritise care. Instead of waiting months for rushed visual checks, people could be scanned at clinics or even pharmacies, with flagged results routed to specialists. It is a practical, scalable way to expand dermatology capacity without replacing clinicians. Jonathan also shares what comes next: millimetre‑wave imaging to look beneath the surface, a human‑in‑the‑loop workflow for safety, and a data strategy that follows regulatory guidance by training on consistently acquired images and biopsy‑confirmed outcomes.

Beyond the tech, we dig into trust. Jonathan outlines a plain‑spoken covenant: transparent consent, meaningful opt‑out, and a firm promise not to sell data. When models improve, retrospective reviews can benefit the same patients who contributed, turning participation into shared progress. He also reflects on leadership in a risky space—hardware, health, regulation—and why a mission that saves lives powers teams through the hard parts. Expect candid insights on aligning patients, clinicians, and payers, managing milestones, and keeping ethics at the core.

Make a small move that matters: book a skin check and talk to your loved ones about doing the same. If you want updates on when you can try the system, join the waitlist at https://www.skinbit.co/. If this conversation resonates, subscribe, share it with a friend, and leave a review to help more people find it.

Chapters

• 0:03 Setting The Ground Rules
• 0:50 Introducing Jonathan Benasaya
• 2:32 The Melanoma Wake-Up Call
• 7:20 From Music Tech To Med Tech
• 11:43 The Dermatology Bottleneck
• 15:13 Building A Full-Body Skin Scanner
• 18:10 Data, Regulation, And Validation
• 20:25 Beyond Cancer: Cosmetic Use Cases
• 21:01 Ethics, Trust, And Patient Consent
• 24:20 Risk, Leadership, And Team Motivation
• 28:20 Final Advice And How To Get Involved
• 30:16 Closing Thanks And Next Steps

Transcript

Stephen King: 0:03

And welcome. We are on a new episode of Incongruent, and this time we are talking health AI in the realm of healthcare and the early detection of skin cancers and melanomas. Obviously, we are not a health podcast. We are not doctors. We are not medically qualified in any sense of the imagination. So this is not any form of advice or recommendations. This is purely a discussion on an AI startup that is looking at helping out in this particular field and what AI could possibly do in this field in the future. I am joined today on this investigation by not Dr. Radhika Mathur.

Radhika Mathur: 0:42

Hi everyone.

Stephen King: 0:44

Good. So Radhika, who do Not Dr. Radhika, how do who did we speak to today?

Radhika Mathur: 0:50

So we spoke to Jen Jonathan Benassaya, and he has had an amazing and inspiring entrepreneurial career as well as personal journey, which led him to Skinbit, which is what we'll be discussing on the podcast today.

Stephen King: 1:06

And he's also been involved with at least two amazing companies that we're both aware of, one of them being Deezer, and the other one being Life360. He's also involved with Meta at some point. And he's really is one of the technology startup Hall to Fame. If there ever was one of the equivalent of the Rolling Stones magazine, Hall of Flame entrepreneur technology startups. He's definitely up there. So we're really honored to be here, right, Radhika?

Radhika Mathur: 1:37

Yes, absolutely. It was such an honor to talk to him. And I cannot wait for every one of you guys to hear what we discussed. So here we go.

Stephen King: 1:49

And welcome to a new episode of The Incongruent. And today we have a very special discussion. We're going to be talking about AI in the context of healthcare. Now, obviously, we are not a healthcare podcast, and we're not making any recommendations to uh use any particular service, or we're not making any kind of uh review of any particular service. This is a discussion on potential applications of AI in healthcare. Uh today joining me is Radhika Mahur. Hello, Radhika.

Radhika Mathur: 2:17

Hi.

Stephen King: 2:18

And Radhika, you're going to be taking over from me here, so let me pass the baton to you and please proceed.

Radhika Mathur: 2:25

So welcome everyone to another episode, and we are proud to welcome Jonathan Benasaya on our podcast today. Jonathan is a serial entrepreneur passionate about improving people's lives through technology. In 2007, he co-founded Deezer. In 2013, he founded StreamNation, and in 2017 he joined Matrix Partners as entrepreneur in residence and partner on FinTech Investments. During COVID-19, he joined Life 360 as his first chief product officer and then Meta to manage their payment platform.

Radhika Mathur: 3:05

Jonathan has was diagnosed with melanoma in 2020 and has been through the traumatic process of dealing with skin cancer. Throughout his treatment course, he experienced and interacted with antiquated and inefficient and an inefficient system. After his treatment, he started to look for solutions to find a better way to prevent, detect, and treat skin cancer, which brought him to create Skinbit. At SkinBit, he leverages breakthrough technology to totally rethink how we detect and address skin cancer.

Radhika Mathur: 3:40

He has developed millimeter wave imaging that identifies cancer in under a minute and with unmatched accuracy. It's a concrete example of using new tech to not only improve but fundamentally shift an industry that's been stuck with slow manual workflows. Welcome to our podcast, Jonathan.

Jonathan Benassaya: 3:59

Hi and thanks for having me.

Radhika Mathur: 4:02

So we'll just get right into it because we're so excited to talk to you about all of this. You've spoken about your melanoma diagnosis as both traumatic and transformative. Was there a specific moment during your treatment when you realized the existing system was broken and that you might be the one to fix it?

Jonathan Benassaya: 4:21

It's it's not only one moment, it's multiple moments. So first my melanoma was missed because um it was during COVID, so we had masks, and the doctor looked at me with her with her eyes, and after two minutes, she said I was fine. And it's only when I was paying my uh my visit at the reception that um I don't know, I maybe I had like a like an angel looking up for me because at that moment the doctor went at the reception desk, like the counter, and she took a file and she looked at me and she was like, Oh, I forgot to check beneath your mask.

Jonathan Benassaya: 5:03

And then I went back to the examination room, and then she looked beneath my mask, and she saw this very like this dark spot that didn't look like different than the others, and she felt like this one needed a biopsy, which I pushed back on because I had many biopsies in my life and they always came back negative, which uh which raised the question of the efficiency of this approach.

Jonathan Benassaya: 5:23

But that's probably the moment where I felt like, wow, could you imagine if I had left that office a little bit faster? Could you imagine? My melonoma would have been probably metastatic when when when the whole the whole scenes would have been diagnosed. So that's one.

Jonathan Benassaya: 5:41

And then second, when I went for my treatment, I was uh I was told by the doctor that uh the standard of care was um what is called a wide margin excision. Long story short, a huge patch of tissue being removed for a very small lesion. Uh, we're talking about 25 millimeter diameter of skin for a three-millimeter lesion, so one to ten ratio. Actually, one to a hundred because we're talking about surfaces. If I did not uh push back on on this very specific treatment, I would have lost uh half my nose. Which um I I actually advocated for myself and eventually prevailed because she only removed 15% of what was supposed to be removed.

Jonathan Benassaya: 6:27

So when I saw that, I was like, holy cow, pardon my French, but holy cow, there is something weird uh about this whole thing, and that's when I started to dig deeper on uh on skin cancer, and um, and that's that's what led me to create SkinBit.

Radhika Mathur: 6:51

Your entire story is just really shocking because, like you said, it could have very easily been missed, it was just like a fraction of a moment, but thankfully it wasn't. And I'll move on to our next question. You helped pioneer digital transformation in entertainment with Deezer, and now in healthcare with Skinbit. What parallels and what profound differences do you see between disrupting how we listen to music and how we detect cancer?

Jonathan Benassaya: 7:20

So if you take a step back, when you look at music, music is um it's like um it's like a tripod. You have uh the content, you have the distribution, and you have the users. Okay, so and you need those three to be able to work. And in health, you have the patients, you have the doctors, and you have the eventually the insurance. So you you still have like three different like types of actors, and you cannot work with only one, like it has to be the three of them. So working in um, you know, usually in business you work with only one counterpart, all right? But in those like very specific type of business, you have to play a dance or you have to dance a dance that requires the to navigate different type of stakeholders with different type of objectives. And if you look at if you look at these are the main challenge was how do we move from an access from an acquisition model, an ownership model, to an access model. Which was what the the consumer wanted, the distribution did not exist, and the labels were like, no way we're selling discs.

Jonathan Benassaya: 8:52

So having to basically align those three types of stakeholders was kind of a very interesting challenge. And um the leverage was more the regulatory framework that basically helped like being a catalyst. I don't know if you remember, but this was founded in France, and and France was the very first country to basically support this model.

Jonathan Benassaya: 9:16

The what basically put everyone around the table and aligned everyone was piracy was rampant and and it was killing this industry. So um, so the French government pushed the older stakeholders to sit at the table, and basically we were advocating for being the solution against piracy, and that's what basically drove the labels to start working with us because they knew that by having this alternative they would be able to work with uh the administration on on piracy. So that's just to give you the framework.

Jonathan Benassaya: 9:51

If you go to health, uh, we're not talking about piracy, we're talking about shortage of doctors, we're talking about doctors that are completely overwhelmed, we're talking about diagnoses that are later and later, uh, which basically creates a spike in skin cancer treatment. So you have a combination of uh challenging access to dermatology plus a very manual uh approach, meaning that whether you or me go to a dermatologist, we they still have to look at us with their eyes. And it's kind of a not very efficient way of doing things.

Jonathan Benassaya: 10:29

You have health systems that are spending billions and billions of dollars, like skin cancer is growing at a the cost of treating skin cancer is growing at more than 16% year over year in the US alone. And you have patients that struggle to see the doctors, they have delay diagnosis, meaning that uh they're they have they need to have like larger incision for uh the treatment, their treatments are more invasive.

Jonathan Benassaya: 10:59

So it's very different than music, but on the other end, it's it's also a dance that you dance at three, and in that case, it's the doctors, the patients, and and the healthcare, the healthcare system. So it's a long story, it's a long answer. Sorry, to to say that uh in uh in Skinbit and in Deezer, we we have to deal with this multifaceted type of uh business dynamics and market dynamics, and and I feel like Deezer teach me how to deal with them, and and that's how I'm able to navigate that uh with Skinbit.

Radhika Mathur: 11:33

That was a really wonderful analogy and comparison that you made across the two sectors despite their differences, which takes us on to our next question. Can you explain what the technology at Skinbit actually does and what makes it revolutionary compared to traditional dermatological tools?

Jonathan Benassaya: 11:53

So, before I answer your question, I think it's very important to understand what's the challenge. All right, and as I said, there is not enough doctors today. If you take your phone and you try to get an appointment with the derm, it will take between three months and 12 months to get an appointment. So that's one.

Jonathan Benassaya: 12:08

Second, doctors, because there is not enough, doctors are completely squeezed, they have to see 40, 50 patients a day. So their performance is not the best. I mean, the doctor you see at 8 a.m. is probably not as good as the same doctor at 5 p.m. That's kind of human. We're all humans.

Jonathan Benassaya: 12:29

So the premise of Skinbit is to basically how can we help doctors to really focus on patients who really need them? And how do we help patients to get an easier access to uh dermatology care? So when you start combining both, you ask yourself, how do you do that? The way to do that is to basically do what we call a triage so that you can get your skin checked before you even see a doctor. And this triage can be performed by a scanner, and I'll tell you more about it.

Jonathan Benassaya: 13:05

A scanner that has an extremely high resolution. We we call it dermatoscopic resolution. It's it's a it's it's kind of a uh dermatologic lingo, but it just means that it's it's kind of magnified at uh the equivalent of a 10x magnification. So we we basically are able to image your entire skin at a resolution that is a clinical resolution, and that allows us to detect all your lesions, understand them over time, if they've changed, if you have new lesions, and for the ones that are kind of weird, allows us to using AI to identify or score their potential malignancy.

Jonathan Benassaya: 13:44

When you're able to do that, then uh you can you can shift the paradigm of like the workflow of a doctor. Instead of having all the patients being seen by the doctor, you would have all the patients being scanned by the scanner, and the the doctor will only see the patients that the scanner has flagged.

Jonathan Benassaya: 14:04

So, on one end, it's way faster for patients to basically get screened, and it's better for doctors because all of a sudden they don't have to see patients who don't really need them, and they can focus on the patient who really need their care. So that's really the I would say the paradigm shift that we are advocating for.

Jonathan Benassaya: 14:23

And in order to achieve that, Radhika, we had to think about how can we create these images? How can we create a digital twin of your skin that has enough information for a dermatologist to trust the system to be as good as them doing the screening with their own eyes? Unfortunately, this technology did not exist in the market.

Jonathan Benassaya: 14:52

There are very few machines that are able to image your skin at a very high resolution in its entirety, and none are close to what is required to get the clinical resolution, and that's what Skinbit has been building.

Jonathan Benassaya: 15:10

So Skinbit has been is building. I actually was in Paris last week to get the delivery of our very first scanner, and and this technology that we've developed is basically a full-body scanner that scans your skin at what we call dermatoscopic resolution, which is the clinical resolution required by a doctor to actually put a diagnosis on something. And as you said during the introduction, we're also working on a second modality, second technology that will allow us to not only image the skin on the surface, but also image the skin under the surface. And it will take time for this technology to be available in the market, but the very first version, which is full-body photography at this clinical resolution, will be deployed next year.

Jonathan Benassaya: 16:02

So that's that's that's what was uh wasn't available in the market and what Skinbitt is building. And by by building this machine that's completely autonomous, that can be deployed at any point of care, you can augment the dermatologic screening capacity of any state, anywhere. The machine has been designed so you can you can have it even in a pharmacy. You don't have to go to have the dermatologist. So all of a sudden, you can not only increase the dermatology capacity of of a country, uh, but you also um help doctors to do a better job and focus on the right people. So that's um that's really the vision behind Skinbit.

Stephen King: 16:43

So once you've developed the camera or the lens or the these imaging technology, you're then gonna need to have data within your model to be able to make a diagnosis now. As you say, it's not possible, or is it possible? So where were you going to get that data from in order that you're gonna be able to make the predictions, or is there gonna be a phase of we're gonna do this service first and then uh this will be later?

Jonathan Benassaya: 17:13

Uh it's a very good question, Stephen. So just so you know where the the regulations is going, whether it's in Europe or in the US, is that it's if you want to apply an AI model, this AI needs to be trained with the data coming from the same technology over and over because they want the those like administrative offices want consistency in acquisition, consistency in treatment, and consistency in training.

Jonathan Benassaya: 17:46

What we've done is we've built a model today based on 50,000 agents. We've acquired those images through a mix of public data and private data from dermatology clinics. But that that is just to basically train and start. It's basically a way to start. Um, and then what we're gonna do is once we deploy our first scanner in Q2 next year, we will basically have a manual review of all our scans with our own dermatologist. And the the scanner will be deployed in a dermatology clinic so that if the machine flags something, a dermatologist can do a biopsy and we will get the biopsy results. And that will allow us to create an entire data set that is 100% proprietary, but also coming from our technology with our own images and with a diagnosis that is confirmed through histopathology.

Jonathan Benassaya: 18:50

So it's a two-step process. One is how do you kickstart or how do you bootstrap, and that's what we've done. And then next year, when we'll have our own machine deployed in the in the field, then we will replace this data set with a data set that is uh entirely built through our own technology.

Stephen King: 19:10

So that's gonna take quite some time, but is there from an entrepreneurial point of view, there's a minimum minimum viable something. Is there uh a non-medical use that people could use this for, or is it is that not desirable because it could create false positive sort of things or something?

Jonathan Benassaya: 19:28

So, first of all, yes, it will take time, but not that much time because each time you go, I mean, humans have on average 100 lesions. Uh so the data set is being built super fast. So we're not we're not really worried about it. And the second thing is uh we've built a workflow internally to be able to um basically run those um review, those screening with doctors in-house.

Jonathan Benassaya: 19:54

So we have before we get to a point where the AI is taking over, we still have some time to get to a few dozens or even hundreds deployments without uh a need for our AI to be um fully uh operational. Now, you're asking a very good question.

Jonathan Benassaya: 20:14

Can 3D reconstruction at high resolution be used in other fields? And um, we have seen a lot of um plastic surgeons coming to us and they were saying, Well, let's flip the script, let's use your scanner to show a before and after treatment. And on top of that, we'll screen for scan for skin cancer. So there is a natural, as I would say, um like cosmetic uh use case uh that uh can also be used for this uh modality.

Stephen King: 20:51

Super. I shall be using it for that purpose as then to get rid of some of my uh under eyes. Uh Radhika, let's move on.

Radhika Mathur: 21:01

So ethics is always a great consideration when it comes to AI. So having worked at Meta and Life360, you've seen both sides of AI, scalable innovation and ethical complexity. How do you ensure that Skinbit's use of AI remains grounded in human well-being rather than data monetization?

Jonathan Benassaya: 21:21

I think that's one of the key concerns, to be honest. As we speak, we are we're working on, like we're gonna publish on our website the Skinbit Covenant, uh, where we uh we basically uh are very uh honest, transparent, and upfront about what we do and what we promise to do and what we promise not to do.

Jonathan Benassaya: 21:42

And on that that basically vein, we are going to focus on transparency with our users. So, first, we want to make sure that they understand what we're doing. Second, we will never use those data like uh to sell uh more of our to sell the data to third parties, so everything that we get in skin beat stays in skin beat.

Jonathan Benassaya: 22:09

And then you need to also let the patients opt out of this whole thing. Honestly, I've seen so many dark patterns where users don't even know that their data are being used to train a model, and that's not what we want. We want people to feel that when they participate in the Skinbit program, they help themselves but also it they help others, and that's that's basically the magic when you work in healthcare.

Jonathan Benassaya: 22:38

It's a little bit different when your data are there to benefit financially to someone else. Here, the data are here to benefit yourself if there are improvements in the model and we can uh retroactively check your photos from the past and see that something was missed. I mean, it benefits you as much as it benefits others.

Jonathan Benassaya: 23:01

I'm gonna repeat what I said. It's about transparency, it's about informing, and it's about letting people opt out. I think that once you do that, you create trust. And once you create trust, that's where where you can build the best things and you you create the best brand. So it's not only about AI, it's also about building a brand and being building trust with uh with your users.

Stephen King: 23:24

We've got so many different speakers echoing that in different fields. I think the first guest we had was talking about vision computing from a security angle and how they wouldn't capture biometric data because they they just did not feel that that was something that they wanted to have on their servers and have all of that difficulties. And then just a couple of days ago, we were uh who were we talking to? We were digital twins of people from their uh public statements, government ministers. So combining technologies together it could create like uh proper full-on digit digital humans. So we're we're looking at maybe in the future an Amazon or uh Microsoft buying different services. Anyway, that's something completely bought in my head.

Jonathan Benassaya: 24:13

I agree with you, and and and I feel like you can build the worst and you can build the best. I feel like our mission is really around detecting skin cancer, and it is so I feel like really humble to be the steward of this venture, but at the end of the day, the objective is to help people's life. So I feel like I feel like the um the mission is so important. I have no problem for people to opt out. That's their own. We won't be able to detect, I mean, to see if we can detect new things for them. But when you talk to people about like what you do with data, what's what's what is it's going to be used for, yeah, they are generally pretty excited about it.

Stephen King: 25:04

Super. Uh super. So going back to entrepreneurship and leadership, okay? There's always in leadership, they say in order to be an entrepreneur, you have to be willing to take risk. And risk from one perspective is a nice concept, but you've gone through an illness uh which has changed your understanding of what what that means. So what do you consider now? Has it changed your experience on what risk and leadership means?

Jonathan Benassaya: 25:40

I think that on the risk side, I mean you can see I'm probably tackling one of the it's probably the riskiest venture of my entire life, to be honest. You would have probably thought that by being a little bit older, I would be more risk-averse. I'm actually increasing my level of risk, but I feel like I've so many scars that I am not going at it completely blind. Uh, I feel like I'm anticipating a lot of things and and I can use my uh my experience to drive the team and and and basically de-risk a lot of the different aspects of it.

Jonathan Benassaya: 26:25

So that's that's really on the risk front. But really to go back to the question, I think it's all about leadership and being able to shift the script by empowering people. I feel like the most important part of my job today is to people feel like they can have an impact on others' life, and and that basically drives their motivation, that drives their ownership and accountability. So on one end, it's extremely risky, but on the other end, the mission resonates so much with the people we work with that I don't have to push them.

Jonathan Benassaya: 27:02

Like there, I'm like, of course, I need to to basically organize and make sure we stay on on track and we maintain the pace and and I I find solutions with them. But for the first time, I feel like the mission resonates so much that the leadership is kind of a it's kind of a natural work, it's really a collaboration. And um, and I'm I'm also sharing with the team where we need to be careful and and where we need to prioritize, because some of them would want to go more in one direction or the other, and and I also have to tell them, hey, what about the next man raising? What is the milestone we need to reach?

Jonathan Benassaya: 27:43

You think that this is going to help? So, you know, I'm I am more on the coach side than I was on uh uh before, and it's uh it's very fun. So it's it's very interesting, Stephen. I I feel like it's a very risky bet in the sense that it there is hardware, it's health, it's a lot of very important things, but it's not that risky when you have the right people and they are motivated by the right things, and and I think this is how you you can build great things.

Stephen King: 28:15

I think you've answered pretty much all of our questions in your conversations, which is amazing. If there is one call to action for everybody out there, what would you want them to do today? Or or if you either from a regulator, you compare from a consumer or from a medical practitioner.

Jonathan Benassaya: 28:35

Is there anything that you would the first thing is I would ask everyone to get their skin checked on a regular basis? So I I'm not here to advocate for my brand or my service, I want people to feel like they have a good handle on their health, and skin health, unfortunately, is very often left on the side, and when it hits you, my friend, it eats you bad, and you can die, you can be dying without knowing.

Jonathan Benassaya: 29:03

So that's the thing. Okay, so do me a favor, even if you're 20-year-old, go see a dermatologist, even if it's every two years, it's totally fine. Okay, so that's really my first call to action.

Jonathan Benassaya: 29:16

The second one is um, if you want to stay informed uh around what we're doing and and when you will be able to test the skin beat system, you can register uh on our wait list on uh on our website, www.skinbeat.co. And that's it, Steven. I cannot ask for more than that. I just want people to be aware of their skin because um again, when it eats you, it's uh it's not as if you were like bleeding, it's not itching, it's not weird, it's just there, it's killing you.

Stephen King: 29:49

Fine. Thank you so much. That's the most jolly ending almost. Uh Radhika, would you like to ask anything else or for this for our wonderful guest, Jonathan? Or do you want to close this?

Radhika Mathur: 30:00

I just wanted to say that it's amazing learning about this new technology, and also I think I'm going to book an appointment with my donate dermatologist ASAP. I'm scared now. But uh no, it's great what you're doing, and we wish Skinbit the best. So I'll just bring us to a close. So thank you to all of our listeners for tuning in, and thank you so much, Jonathan, for being on our podcast. Everybody like, share, comment wherever possible, and please be sure to tune in to our next episode. Bye.

Jonathan Benassaya: 30:32

Thank you.

Editor: 30:33

 Last edited, December 31, 2025.