[00:00:00] Luis: It you don't fall in love with a problem, and instead you fall in love with a product, you're doomed. And I think that was something that really helped us fundraise . I'm not sure what investors think of me. I think they have a very big certainty that I am very passionate about solving a particular problem. And yeah, that helps build conviction that even if times get hard and if we can't raise or we can't hire or we can't solve a particular tech problem, the founder will just keep going.
[00:00:27] Michel: Hi everyone and welcome to Growth Leap. I'm your host, Michel Gagnon. We talked to pretty awesome business builders who are designing disruptive and meaningful companies.
[00:00:40] Hi everyone, and welcome. Today we're in for a treat as we sit down with Luis Valente, the CEO and co-founder of iLoF, a deep tech startup focusing on personalized medicine. Luis isn't just any entrepreneur. He's a Forbes 30 under 30 awardee in science and healthcare with a background in computer science and a heart fully committed to personalized medicine. Can you believe he started his entrepreneurial journey at just 18? Now he's at the forefront of iLoF leading the charge in using AI to transform how we approach complex diseases. In this conversation, we'll get a personal look at Luis journey, his passion for integrated technology with healthcare and his ambitious dreams for future, where personalized treatments are the norm.
[00:01:27]
[00:01:31] Michel: Luis, thank you so much for joining us.
[00:01:34] Understanding iLoF and Personalized Medicine
[00:01:34] Michel: To get started, can you tell us a bit about who you are and what problem you are solving with iLoF?
[00:01:40] Luis: Absolutely. Michel and thank you so much for yeah. Inviting me. So I am, again, Luis, I'm CEO of file off and at of we are basically collecting masses amounts of data on biological profiles to help, uh, pharmaceutical companies and fundamentally researchers and biotechs, research institutes and, and pharmacists to develop personalized medicines. Quicker, faster, and while creating a much more humane and convenient experience for patients on these clinical studies. we're a 30-person team and we have a very, very strong passion about making personalized medicine a reality rather than just something that it appears to be coming in the very distant future.
[00:02:24] Data and iLoF's Approach to Personalized Medicine
[00:02:24] Michel: Wonderful. And can you explain to me and our audience, a little bit what is it that you do? Because it's relatively complex, especially for those, of us who are not necessarily, uh, you know, well-versed in healthcare and technology. So a quick, high level understanding of what is it that you do?
[00:02:41] Luis: Excellent question. actually, when we hire people, we normally say that we have two dogmas. We have two things that we fundamentally believe with. The first one is that personalized medicine is a question of, when and not a fifth. So it's just a matter of time. And the second is that reason why personalized medicine is not a reality today.
[00:03:00] The reason why potentially you and me, Michel, probably made only a handful if none, genomics or like personalized, tests, it's mostly a data issue. So in order for personalized medicine to be a reality, you need more data. You need better ways to acquire it. You need, better ways to process it. And this is fundamentally the issue that we've been solving. So what we do is we integrate, data that gives us information on biological profiles, and we integrate it from multiple sources all the way from just data on comorbidities, age, sex data that is usually found on electronic health records.
[00:03:38] And then we augmented using multiple mixed data with a very big focus on optical data or optomics data. And this is data that is extracted from biological samples using light. So for the ones familiar with a Nobel Prize for physics this year. You'll be familiar with this approach, but basically what we do is we use, light to, understand the composition of patients, biological samples.
[00:04:04] And then, we train artificial intelligence of machine learning and deep learning models to understand this heterogeneity, understand this composition, and then link it to either the way a patient reacts to a certain disease. Or the way a certain a patient reacts to a certain treatment. So basically, based on this ensemble of information and this ensemble of particles in patient's biological fluids, we are able to say if a patient is going to react very well to a certain treatment or not. And also if the patient is going to react in a very aggressive way to a certain disease, or if it's just going to have a mild or moderate course of that disease. And we've done it for a couple of specific indications, including Alzheimer's, ovarian cancer, and even on Covid, with an interesting case study on that. But basically what we do again is we collect this datasets, we extract valuable information, we feed artificial intelligence models, and then we help at this point, scientists and researchers select the right patients for their clinical studies and for, the right patient is to each treatment with a view of one they help also, medical doctors make decisions about the right treatment to each patient in a clinical setting.
[00:05:18] Michel: Okay, so you kind of touched a little bit on my next question.
[00:05:21] 400 Failed Clinical Trials and the Importance of Patient Selection
[00:05:21] Michel: I was trying to understand where you fit a little bit in the supply chain or the complex world of healthcare. So basically your main stakeholders or client would be medical researchers or, you know bigger companies.
[00:05:33] Luis: yeah. It would be researchers and pharmas and biotechs. And I'll give you a clear example why we've made that strategic decisions. So just for the last century, more than, 400 clinical trials failed while trying to get an effective treatment for Alzheimer's to the market. So 400 clinical trials failed that basically didn't went anywhere, with a massive cost for well biotechs and pharmas and also a cost for patients that are currently living an in the Alzheimer's field. I'm sure. we're all familiar with how big of a problem it is, and with millions of patients that are still living without an effective treatment to that disease. and the reason for that is that again, it, Alzheimer's, there's a lot of others, diseases are very heterogeneous. The way that a patient reacts to a certain treatment is going to be very different and yeah, basically we understood that there was a big problem to be solved in the drug discovery and drug development space. We need to first help these researchers, this is biotechs, get their effective medication to the market faster. In order to do it, we need to help them. Of course, first of all, select the right patient to the study, but we also need to improve the patient experience in these studies. To give you an example, if you wanna participate in an Alzheimer's clinical trial today, you need to do a lumbar puncture, which involves seeking a needle, this back, so 10 centimeter needle into your back, which I mean most people don't wanna do, and especially if you're already in the fragile situation, and this creates huge barriers to recruiting patients to these trials. 90% of them drop out and you can't get these treatments into the market unless you test them first. So there's a huge bottleneck into getting these personalized treatments into the market. And this is the problem that we are first and foremost solving by working with farmers and biotechs. Of course. When personalized medicine and personalized therapeutics are abundant in the market, we'll be in a unique position to then be the go-to platform to help clinicians make decisions about which one is the right treatment to the patient but right now, we're still not there yet and that's why our main counterparts are farmers and biotechs.
[00:07:41] Apply the Netflix' Personalization Model to Clinical Trial
[00:07:41] Michel: So if I understand correctly, you're simplifying the tests themselves, or how you select patients and by doing that you have, let's say a broader pool of candidates and then you can, segment them, you know, a little bit like would Netflix would do with us. Uh, yeah. And by doing this, well, you increase the pool of candidates. You also increase the chances of results in terms of tests, that kind of stuff.
[00:08:08] Luis: Absolutely. And you've just spoken about Netflix. I. So the way that Netflix, no, no. It's an excellent example. The way that Netflix work and the way that Netflix gets you hooked is that it hopefully learns enough about you to give you recommendations about movies that you'd even even knew you wanna see and that you would have no way or knowing if you wanna see or not. If Netflix hasn't evolved into such a company and especially since there was a strategic decision to more and more remove from the Netflix catalog, the blockbusters the movies that everyone knows about. So it's even more reliant on its ability to deliver you unknown gems and like very, very good movies that you dunno about.
[00:08:51] Right? Uh, the thing is that in medicine. When you go to the doctor for a lot of diseases, the clinician will go to you and will say, Michel I have no idea if this treatment is going to work for you or not? We'll try it out. Let's try out this one and then if it doesn't work, we'll come back and we'll try another one. And this is just madness. And it's not just on Netflix that there's, so there's an even more clear example when you go to a clothes store and you wanna buy a T-shirt. Right? You're not like, okay, bring me all the sizes and then I'll see if it fits. No, I mean, you have a very good idea of what is the right size for you. Right. You know okay. Maybe you bring me like the M and the L and then we'll see if, I mean, but it's either one or the other. I know because I already know my size. And that doesn't happen in healthcare. So just imagine if you were to a clothes shop and you tried like a size randomly and try to put it on and you say, oh, this shirt is bad, this shirt doesn't work it's madness, right? It works. It's just you're not the right fit for it. And this happens in clinical trials all the time, which is it in a lot of cases it's not that the medication doesn't work, it's just you're trying it on the wrong people and because you're trying it on the wrong people you have this impression that it doesn't work because you're trying one person and then it works. And then on the other person it has no effect. And then on the third person, it can actually kill that person or have like very, very adverse side effects. And if you could only test that medication on the kind of person, that phenotype of people that work, that have the right biological profile, then you could get a medication that doesn't work for everyone. And this is the basis of personalized medicine. It's the one will not fit all approach, but at the same time, you'd get an effective treatment to the market that will make an impact on the lives of millions of patients. It just wouldn't be the entire patient population with that disease, but it'd still be hopefully a very large set. And yeah, this is basically what we're setting out to do, where we wanna give clinical trials, more tools to select so if they're selling XL t-shirts, then we just wanted them to try out those T-shirts in XL customers. And not just the whole array of people because if we test them on the whole array of people, then most people will say, this shirt is no good. Doesn't work for me. And we try to do that with treatments.
[00:11:03] The Only Reason You Should Start a Deep Tech Company According to Luis
[00:11:03] Michel: I wanna talk about your journey as an entrepreneur. But before we get there, I want to really talk about how iLoF started. In today's world where, you know, everything is instantaneous, uh, we see a lot of of people on on social media saying if you want to get rich, you just need an iPhone and a credit card and, you know, make a lot of content.
[00:11:25] When you think about a deep tech company, it's, it's not as simple as that. And, uh, we hear a lot of, um, advisors or very well-known entrepreneurs saying the best thing that you can do is to find a very big problem to solve and, you know, work on it and build a business around that, which I believe is what you're doing.
[00:11:44] However, you know, big problems are, are tough to solve. So I'm interested in hearing your story or how you started this. Where do you start? It's not, it's not like you have to, you set up a website, you build a marketing funnel, and then you, you know, you start making cold calls. Right? How, how did you approach that? How and how did that evolve?
[00:12:04] Luis: That's an excellent question and first of all, I have a huge respect for entrepreneurs that made them, that built amazing businesses such as Stripe, for example, that, I mean, empowers, well, hundreds of thousands of small businesses to get payments, right. And other businesses like, I mean other sorts of businesses that we interact on a day-to-day basis, and that were able to create usually successful businesses out of basically using technological tools, but let's say, uh, not deep technological tools, um, to make amazing business.
[00:12:38] I have a huge respect for them. If we think about where we were three years ago, two years ago, during Covid, it wasn't Facebook that saved the day. It was an Instagram that saved the day. It was, it were, it was deep techs, right? With the mRNA vaccines and scientists all across the globe, like pulling their resources and just working together to try to solve a global health situation, right? And if you see some of the biggest advancements in, in the way we work and in the way we live, were made by fundamentally, but we're unlocked by fundamental, complicated technological and scientific problems. Right? And I'm not just speaking about healthcare, so I guess in healthcare, I guess it's obvious, right? But even in our lives, there were technical problems that had to be solved to make the internet work. Or to make, uh, well, God knows rockets work and put satellites in orbit and uh, and again, we don't all need to be rocket scientists because not everything is rocket science, I guess. But at the same time, um, I think deep tech is potentially is going to be the next big thing. And it already is for the next a hundred years. And this is because we have a bigger pool of specialized talent than we ever have. This is because we're quickly getting to a point with, well, with the growth of AI tools that can basically perform more simplified, uh, um, tasks.
[00:14:14] I think more and more humans will converge into solving big, deep tech problems that machines and simpler or dumber algorithm still can't. So that's one. Now the problem is, and there's a, so maybe, maybe you saw, um, recently the Nvidia Co-founder, uh, Jason, uh, Jensen, Jensen Huang, right? And they, and they asked him, okay, what would you do differently if you were to start Nvidia all over again?
[00:14:42] And he was like, I wouldn't, I would never, crazy because it's damn hard. It is hard. So I, I was in a, I was speaking in a conference the other day and the, the room was full of PhD or postdoc students. And they were like, yeah, I mean we're, we're, we just finished our PhD. We don't wanna, don't know what we wanna do.
[00:15:01] Maybe we wanna do a postdoc, maybe we wanna go to a corporate or maybe, I don't know, maybe we'll start a company when we'll raise some, we'll raise some funding . And my answer was like, just absolutely do not start a company in Deep Tech unless that's the only option that it's viable for you and only option, not because it's going to make you money, but only option because you feel compelled to solve a certain problem.
[00:15:25] You can't sleep, you can't just phantom the idea of a certain problem being there and you're not solving it. And you feel compelled to do it if you don't fundamentally feel compelled to do it. If you can imagine any other path where you just do it, because this is damn hard. It's not half as sexy as it seems to be. And you'll, you'll give up along the way. Because it's just, there's a lot of sleepless nights. It's a lot of weekends. Um, there's a lot of things that you hit your head against the wall until it works, and then it works for a while, and then it stops working, and they need to continue hitting your head against the wall, and it's the adventure of a lifetime. I have a peer to peer, like CEO group of Deep Tech co-founders, and I think we all love what we do, but at the same time, we all say just don't, don't do this unless this is the absolute only option for you. Uh, and you can't see you yourself doing anything else because it's very, very, very hard.
[00:16:18] How Luis Started a Deep Tech Startup
[00:16:18] Michel: I do understand the, the challenge. I think even when you work in, just in a digital world, uh, without all of the, the challenges that you have, it's still a, a pretty intense roller coaster. How did you start, how do you start to put the pieces together. Do you, just put a pitch together and, , go see some, VCs or, or, strategic partners. Do you start building something? How did you get started?
[00:16:45] Luis: So I was working in a, in an RD institute and um, I was kind of helping with tech transfer and so on and so forth. And I came across, and this is when, me and the founding team, came across, the opportunity to build tools that could help phenotype patients. And it started out like a very, I mean, we're, we've made a thousand different pivots since then.
[00:17:11] We're not even using the same tech anymore. But it started out as, oh, wouldn't it be cool if we had like a way of using light to collect information about patients' profiles and it, because light is cheap, light is like just light easily. So you can't collect, then you can digitalize light, you can store it on the cloud, you can create a library. It's amazing, right? It's just transformational. But I have to say that I was like, yeah, I mean, this seems amazing, but, um, I just had an offer to join an American company in Mountain View, amazing company. So I, it's an offer I can't refuse. I'll just go. And I left, but while working on that company, which was also an amazing company and also Deep Tech, uh, doing amazing software for car manufacturers to make cars speak between themselves and, and, and just accelerate the next level of mobility. Um, but while working on that company. I started to get very close to, to the project and to the team that was developing it, uh, again, and eventually again, I realized two things. So first of all, I realized that I cared too deeply about the personalized medicine problem. I cared too deeply about the fact that we still fundamentally don't have ways to know if the right treatment approach that we're using, if the treatment approach we're using is the right one. And that was because it was a mix of personal experiences with, uh, with just overall realizing that this is a problem that's potentially holding back millions of people all across the globe. And what if, what a big impact we could have on the world if we were to be the ones to, to contribute to changing that. So that was one. And then second, I started to understand that there was a big billion dollar opportunity to solve it. So I mentioned to you that 400 clinical trials failed in the last, couple of decades, uh, on Alzheimer's alone. Each one of these trials costed on average 5 billion.
[00:19:16] Of course it was spread out along decades and decades, but it's just a massive amount of money, resources, and time that was spent and that is still spent on trying to get these medications into the market and, um, and they just don't work.
[00:19:31] And then you need to start all over again and it doesn't work. You need to start all over again. And this is fundamentally because again, you like the tools to select the right patient to each medication.
[00:19:41] How Luis Did 10 Interviews a Week for 10 Weeks to Validate a Problem at the Harvard/MIT Accelerator
[00:19:41] Luis: So what we've done is , I ended up quitting my job because again, I was compelled to do it and we joined a, uh, an accelerator. So a Boston-based accelerator called, Crash that it's organized by a couple of, um, institutions around Boston, um, Harvard Medical School, MIT couple others. And the objective is to put to test, like literally to crash, uh, and to try and destroy ideas that have an expectation of being able to transform healthcare. And in 10 weeks we were forced to speak with over a whole, over a hundred global stakeholders, key opinion leaders, pharma executives, CEOs of big biotechs to validate number one, the problem. Number two, the approach, and then number three, the the product. So. First the problem, then the approach, and then only then the product.
[00:20:37] Right. And, and it was very interesting because if we, if we didn't do 10 interviews a week, so it's 10 weeks, a hundred interviews documented, we need to write all of them down. If we failed on a single week, and then may, maybe, let's say we've done nine interviews instead of 10, we will be kicked out of the program. Just kicked out. Okay, you're out and, and by the end of this 10 weeks, so first of all, we really like Super Accelerator. We're like, okay, let's do this. We're super hyped. But at the end of the 10 weeks, more important than that, we had 10 letters of intent, three of them from global pharmaceutical businesses that said, number one, this is a huge problem. You couldn't find a bigger problem to solve. Number two, the approach is feasible. And number three, if you ever build a product, we would like to try it and we would love to help you build it. And then with those 10 letters of intent, this is how, this is how we then went to funding, and that's how we eventually raise our first 2 million.
[00:21:38] Michel: Amazing story. Yeah. I like the 10 interview week or you get kicked out. It's pretty straightforward.
[00:21:43] How to Position Your Startup in a Highly Regulated Industry
[00:21:43] Michel: I wanna keep digging on this and, maybe you'll tell me you, you answered the question already. But after that, you operate in a highly regulated space where regulations vary, uh, significantly from one country to another. . It's a space that's dominated by huge conglomerates.
[00:22:01] I'm curious to know how you go to market with a company like yours, right? So you've kind of talked about the inception to a, a certain extent, but, once you have that set up, what are the next steps and, how do you win more clients? What's the approach?
[00:22:17] Luis: So one of the reasons we chose this current Go-to-market strategy of going to Pharmac and biotechs was also of course, because the problem was there. That was just elaborating. But at the same time, it also helps that while we're working on the drug discovery and drug development space, there's fundamentally no government regulatory path because you're under the research use only exception. And that what it makes you is that it allows you to create impact without necessarily having to jump through all the hoops of the regulatory path. Now having that said, we are doing both in Europe and in the States with the FDA, a formal regulatory path, to make sure that we achieve C mark in Europe and the FDA approval in the US for the products we're building because as I mentioned in the beginning, we see a clear view of how we can then translate this platform into a clinical decision making tool platform and I can absolutely walk you through the experience and what has it been and what are the challenges. But again, we don't necessarily have to do it, to be able to generate impact and even revenue because we're under the research tools only exception. Now, I can tell you one or two things about the differences between, different geographies.
[00:23:36] And this is a pain. I think this is one of the things that is holding, that whole industry back. And I feel like especially in Europe, we keep following the same regulatory kind of mistake, which is we try to solve everything with just more regulation. Sometimes it helps, but not always. sometimes just regulating your way out of the problem is not going to help. And I can tell you that right now in Healthtech, there's a lot of very, very good companies that are being compelled to go to the US not just because that's where a lot of the money is. But because it, there's an easier and more straightforward, I wouldn't say easier, but more straightforward regulatory path. It's a bit hard to interact with regulatory bodies in Europe. It's a bit easier to do it in the US. The FDA is a lot more approachable. But having that said, I mean, and until we start operating on a global scale as a species, right, and still we're still separated by borders, I think this is the kind of things that people deal with on regulated, on, I wouldn't even say regulated sectors I would say sectors that really matter a lot to a lot of people. And I think it comes with being a health tech founder. You need to adapt. You need to understand that fundamentally there's going to be a lot of overhead trying to adapt your products, to feed the different regulatory constraints. I tend to see regulation as good, except when it's excessive and then it's just dumb. And yeah, I mean it comes with the job description of being a health tech founder. But again, having that said, it's sad that a lot of good European founders are being, pushed to the US not because they don't have the money to build, but because it's just easier to get products on the market in the US.
[00:25:21] iLof's Business Model and 'Foot on the Door' Strategy
[00:25:21] Michel: So you have the opportunity to significantly reduce the cost of drug trials and, and tests what's the business model for you? I've mentioned that you have some, um, and we can talk about your investors. I think, you know, Microsoft is one of the names. Maybe you had one of their folks on your board call earlier today. But, um, what's a bit the business model? How do you make money? Who are your clients and what's the path to hopefully a profitability or an exit?
[00:25:51] Luis: Absolutely. There's a fundamental advantage in the way that we do things, which is we are collecting very large amounts of data, but we are fundamentally collecting data that is easier and, cheaper to collect. It's very noisy data and that's also why it hasn't been collected so far. But it's very cheap to collect. And again, we use light and all of that, so it's very cheap. And this allows us to basically implement since day one, a strategy which we call like a foot on the door kind of strategy where we go to biotechs for example.
[00:26:25] We go kind give you an example of a biotech in the UK we've been working where we go to them and say, okay. Or maybe they go to come to us, which is unfortunately what happens most times because we have still have a small 30-person team in our business development efforts are still very limited. So they come to us and, Hey, I have this problem. I'm trying to take this target to later clinical stages, and I really need help with this, right? And then what we try to do is we try to establish a model where we charge a small fixed fee recurring fixed fee for using the platform, and then a variable fee where the biggest chunk of the money is on the variable fee, that it's directly correlated to the savings we enable the customer. So as a rule of thumb, you can say that we keep around 30% of the savings that we enable a customer to have
[00:27:18] The Harduous path for Alzheimer's Patient Selection in Trials
[00:27:18] Luis: and again, giving the example of Alzheimer's, if you consider that each patient that is currently tested for Alzheimer's a pharma pays around $4,000 to test a patient. 90% of them will not have the right fit, so they're spending $4,000 per patient, but 90% of them don't have the right fit. So they're discarded using invasive methods, and these same invasive methods need to be repeated over time, which leads another 90 of the 10% that have a right fit. 90% of the ones that have the right fit they just, they drop out of the trial because it's a very invasive and inconvenient procedure. So in order right now, and this is, I mean, there's a couple of pharma that speak about this openly because the big problem, I think there's even a, a gates note from Bill Gates around this, especially about Alzheimer's. So in order to recruit one person, you need to start with a hundred because 90% of them will have the right fit. Again, personalized medicine, one does not fit all. And of the 10 that stay, nine, nine will go away because they don't want to put a needle this big into their back every three months. Other reasons inconvenience just it's not a very convenient experience. So our approach is we wanna come in the beginning of the funnel. Okay? And first of all, we wanna select using again, much more. Convenient methods using data that maybe they're already collecting for other sources, for other just routine tests. And then we train our models. We deploy this model in the triage stage, and then we only let path through our filter the patients that have the right fit. What is due is that right off the bat, it saves a bunch of money to pharma that they would be spending on people that don't have a fit anyway. And it cuts a lot of time. And time in pharma is money because the most valuable asset that pharma have is the license for that specific drug and the license, like it's a ticking clock that every day like TikTok. So they need to get the drug into patient's hands faster, not just because of social impact, but also because of financial reasons. And again, all of this generates significant savings to pharma and we can, as a rule of thumb, we can get a kind of a 30% chunk of those savings. And this is the model. why is this a good model? Because it's very, very risk-free for pharma. we get data which further reinforces our models and it helps us get more impact. More impact regenerates, more case studies, more case studies generate. More partners. More partners generate more data, and then it's a positive feedback loop.
[00:30:04] Michel: So you, get you charge a fixed fee for people to use the platform.
[00:30:09] Luis: Yeah.
[00:30:10] Michel: Then you get 30% of the savings and the way you, can put Rule of thumb, but the way you can put a price on the savings is that, you know, you already know how much it costs per you know,
[00:30:21] Luis: Yeah. That they have. So there's one thing that pharma are good is with the data. So they understand exactly the cost structure and they can estimate how much they budgeted for a certain part of the trial. Whatever the amount is that they save, we get a cut.
[00:30:36] How Successful is iLof's Technology
[00:30:36] Michel: And how, um, successful have you been, with, you know, your model and processing the data? Because I mean, conceptually it makes all, it makes perfect sense, but you need to pump out the right segmentations of the right type of people.
[00:30:50] Luis: so we have absolutely, we have a couple of papers out. And, it kind of depends also on the application. but historically we've been able to show, again, using an average, I would say we've been able to prove concordance to the gold standards around. With a raw KOC of around 80%. some cases actually a lot higher and some cases slightly lower. But it depends on the application. It depends on the profile that we're looking, but again, 80% of the times we've shown in cases where 80% of the times we can proof concordance to gold standards and that basically, uh, well, it's still not a hundred, which we would love to be but it, it's sufficient to potentially save a lot of time and money to these trials.
[00:31:31] And again, as we collect more data, we also refine our models and we're seeing, clearly seeing this upwards trend in the performance of our models.
[00:31:39] michel_1_12-20-2023_171358: Great. I don't want to take too much of your time, but I got a few more questions.
[00:31:43] What Keeps Luis Awake at Night
[00:31:43] michel_1_12-20-2023_171358: One of them is, you mentioned earlier that it's pretty intense, you know, to be, uh, launching a deep tech startup and you should only do it if you feel compelled and you have that urge to really work on that problem.
[00:31:56] What, what keeps you up at night these days? Well, what are the, you know, the top challenges that you're like, oh my God, if we can solve this, that'd be amazing.
[00:32:05] Luis: Well, number one is, so deep tech companies leave out of their people. it's just people. It's not like I can get grab like random people on the street and teach them, physics. I mean, I can't, it's just going to take me 10 years. But, and we don't have that time, right? so hiding. Hiding is certainly, and we're, we're sourcing globally. We're sourcing globally. I was doing a fun, analysis the other day that I, I think in the last weeks or so, I've spoken with people from 25 different countries as candidates. We're sourcing globally because talent is everywhere and we need to get the best of the best, and, And I do believe we have also a very compelling mission and a very compelling company.
[00:32:46] So that also allows us to, bring the best, but we need to find them. Right. And the sad part is that unfortunately, there's a good amount of people that either because they're very successful or very good, or they just built a very big network that they never apply to companies. And especially as a startup, I mean, if we were Google that has Google has like 150,000 candidates a year, right? I'm sure in those 150,000 candidates, there's a bunch of like stellar people. But as a startup, you number one, you can't compete with the brand recognition of Google. Number two, you can't compete with the salaries of Google. So we do a lot of outbound, we just reach out to people a lot and we wanna, I. I mean, sometimes we poach people from like amazing, like companies where they then join us on a lot more, let's say, affordable compensation, packages because they really care about the mission.
[00:33:38] But again, we need to find those people. And hiding is one of the things that keeps me up, uh, awake at night, literally, because I literally sometimes have interviews in the middle of the night because there's someone in Australia or New Zealand or whatever. Just, I need to speak with 'em. so it literally keeps me woke up, keeps me up awake at night.
[00:33:54] The second is that, again, deep tech is really hard in the sense that when you're building software, so again, my background is in computer sciences. So when you're building software, you're building on top of things that were built by humans. Okay, so you're working on a computer. A computer was built by humans.
[00:34:11] You're writing a computer code. That computer code was built by a human right. So if you wanna take it to the extreme, everything that you code was developed to be easy and to make sense. Like a compiler, you're compiling your code. Someone had to build that compiler and that someone was a, was a human right.
[00:34:30] So we built it to work. The problem with deep tech is that most times you're working with things that we didn't create. We didn't create the laws of physics, we didn't create biology. We didn't we're working with things that, well, either you're religious or not. If you're religious, you can say, okay, God created it.
[00:34:45] But anyway, it wasn't us one way or another. It wasn't us. So you're dealing with a lot of things that they weren't meant to be. Easy. They weren't meant to be user friendly. They just weren't meant to work in the ways that we wanna work with them. And that creates fundamental, basic problems that you don't have when you're working with software because again, software was already developed to be user friendly and to be used by humans. And that means that it's sometimes hard to plan. Because you just keep finding, the more you unveil the unknown, the more you find problems that you had no idea they were there. and then you say, okay, next week we're going to ship this. And then suddenly, boom, you just uncovered the source of noise.
[00:35:27] They had no idea. Was there for a whole time. And sometimes it's transformational. Sometimes it's really like, whoa, okay, we can clearly see. How good we are versus where we were, how crappy we were 12 months ago. And sometimes it's just, oh God. Like how do I deal with this? This is slowing me down. I just wanna get done with it and just move forward.
[00:35:46] Again, already it just, so I, I was speaking with an entrepreneur the other day and he was telling me a story about a vc. So he was, he was fundraising and he was telling me a story about a vc. And he was also, he's also doing deep tech and, and hardware on top of all things, which. I mean significantly harder. And he was talking with vc and it's like the VC was like, yeah, you know what, John, I can't invest in you. It was like, why not? I'm, I'm not bold enough to do deep tech. Sorry. I'm really, and it's a matter of being, I, I just thought it was very interesting because it's really about being bold and a lot of VCs recognize that they're not bold enough to the deep tech, even if it's one of the most important things of our times.
[00:36:25] So one thing that keeps me awake at night is that I lie. I'm not sure if I'm bold enough to do it, but I try to do it every day and it's just damn hard.
[00:36:33]
[00:36:33] Fundraising and Investor Relations
[00:36:33] Michel: Very good segue into the next question, which is, fundraising and investors. Can you tell us a little bit about, the investors that you have on board, and also maybe you know, how you, you got them on board.
[00:36:46] Luis: Yes. So we operate in the intersection between artificial intelligence or well machine learning, biology and physics. And our investor base is kind of a reflection of this. So we have more tech investors or AI investors like Microsoft. Which has been an amazing partner on this journey. We have more physics based investors, like a sram, which is a big corporate, 60,000 employees in the entire world, based in Germany and Austria actually.
[00:37:16] And then we have , more life science investors. Right. And this is actually one of the challenges of the company because, We are putting people together, not just people like team, but even investors together that wouldn't necessarily sit on the same cap table. Like let's say a physicist wouldn't necessarily work with a biologist.
[00:37:35] A biologist wouldn't necessarily have to work with a computer scientist. And the computer scientist doesn't normally go for coffee with a physicist. And this is a problem at the team level, right? Because we need to get everyone to speak at the same language, but it's even a bigger problem at the board level, right? Because, they're fundamentally having different drives. And, uh, well, but that's also the problem of deep tech. Again, most times you need to be working with multidisciplinary teams aside, which is different from software where you can have a team with everyone has the same background. We're all software engineers.
[00:38:04] Maybe you have one or two product designers or whatever, but that's it. So how did we came to be? I think so again we've captured around 10 million euros so far in our journey, which it seems like a lot, but it's actually not, uh, considering the amount of, well, how much, so how long we've came and the amount of, let's say work that was put into, considering that, again, it's quite a, we need quite, I mean, specialized people. That's actually not a lot, but the way that we were able to do it was really through making sure that we understand very well what big problem we're solving. So I normally start to say, I normally say this saying where It you don't fall in love with a problem, and instead you fall in love with a product, you're doomed. And I think that was something that really helped us fundraise, I think. I'm not sure what investors think of me. I mean, I hope, I hope I know, but I'm not sure. Right. But if there's one thing that I'm sure is that they know that I'm really, really passionate about the problem I think they have a very big certainty that I am very passionate about solving a particular problem. And yeah, that helps build conviction that even if times get hard and if we can't raise or we can't hire or we can't solve a particular tech problem, the founder will just keep going. And yeah, then it's brute force. I mean, I usually say that fundraising is a bit like dating. What is the likelihood of your finding a soulmate when you do one date? Maybe not a lot. What about if you do 10 dates with 10 different people? Maybe it's a bit higher, but still. We had to do well in our last round, 150 dates.
[00:39:43] Actually, to give a very good example, we had. More than a hundred. We had two meetings with, uh, at least two meetings with more than 150 investors. And out of those, we had the lack and the honor of having the round then oversubscribed. And we had to actually do the hard decision of keeping some investors out. But, I have to say that if we hadn't spoken with this 150, we would probably have failed and we would be dead by now and all be doing different things because it's a numbers game and it took us a while until we got the first Yes.
[00:40:17] The moment you get the first yes, then it's easier. Right? Then you quickly go from, oh God, I'm going to die. There's no money to, uh, a no subscribed round. but yeah it, you need to get to that first. Yes. First.
[00:40:30] Michel: My last question before we go.
[00:40:32] Future Plans and Goals for the Next 12 Months
[00:40:32] Michel: I'm interesting in knowing a little bit what's on your plate or what are your plans for the next six to 12 months for iLoF? Like, what can we expect? What would you like to, if we fast forward 12 months and we have another call, like what would you be proud of telling me we've ticked the boxes.
[00:40:50] Luis: Absolutely. So we'll keep hiding, uh, hiding. I would say it's, again, one of the things that keeps me woke awake at night. We're growing the team. We're hiring people all across from biologists, physicists, computer scientists. so again, if you're listening and you want to just comment, solve one of the biggest problems of mankind, just please, uh, gimme a shout on super active on LinkedIn. That's a box that we need to tick. Then the second one is data. We're running, which is potentially the biggest data collection study currently running in Europe in Neurogenerative that is not being run by pharma with more than 12 hospitals, and hopefully around 2000 patients when everything is said and done.
[00:41:30] So getting data to feed into our models to understand patient variability, heterogeneity, to make sure that we're building our models right. That's big that's one of the things that we're doing. And there's a couple of more technical milestones around these data collection studies that we hope to hit in 2024. And then, the third one is that this will be a year where we'll finally have a little bit more bandwidth to go outbound regarding commercial opportunities. So, yeah, we're hiding a, more commercially oriented profiles and we hope to grow, let's say the product and business development team outside of just the founders.
[00:42:06] And that's going to have, of course, a clear translation into the impact we make in the world.
[00:42:10] Michel: Thank you so much for your time. Thanks for squeezing me in right after your board call, and I wish you all the, I wish you all the best. Thanks Luis.
[00:42:19] Luis: My pleasure. Thank you so much.
[00:42:21] Michel: Thanks again for listening, I hope you enjoyed the show. Make sure you subscribe to the podcast. And as usual you can find the show notes at stunandawecom.
[00:42:30]
Business and tech news in 5 minutes or less
100% free. We don’t spam. Unsubscribe whenever.
We're a Growth and Product Design company whose mission is to transform businesses into designers of positive change. We offer learning experiences with online courses and coaching, do-for-you growth services, and a community of people who care and share about building growth systems and user-centric products and creating sustainable businesses.
.png)
.png)
