Latest / Powering the Future Within / Episode 8: Separating What's Buzzy from Bankable with the First AI-Focused Healthcare VC
Transcript
- speaker-0: Welcome to Powering the Future Within, the podcast where we explore the breakthroughs shaping the future of med tech. I'm Omari Bachnit, CEO of ResonantLink Medical. Each episode, we speak with a leader who is driving the next wave of medical innovation, advancing technologies that are smarter, less invasive, and more connected than ever before. This is where the future of health takes shape. Let's dive in. Today's guest is Armin Vidian, co-founder and managing partner at Recode Ventures, a venture firm focused on backing deep technology platforms, transforming healthcare and the life sciences. Armin has spent more than two decades at the intersection of med tech, entrepreneurship, and venture capital, helping build and invest in companies applying advanced technologies like AI, robotics, and data science. Armin. Welcome to Powering the Future with us. speaker-1: Thank you for having me today. speaker-0: Really excited to have you on. So you've managed some significant transitions going from operating and product driven med tech to digital driven med tech and now leading a health AI venture fund. What has that journey been like for you? speaker-1: So it has been an adventure for my curiosity ⁓ as I've moved from tech to medtech to venture to starting a company to venture capital and looking at how technology is changing all kinds of industries, specifically in medtech, pharmaceuticals and other healthcare companies. ⁓ But it has been an adventure focusing on customers, thinking of how technology changes things, and a lot of adjustment and being nimble. speaker-0: Yeah, yeah. And I'm a big believer that a lot of the insights that you have in some of your experiences you bring with you every step of the way. And I'm curious what insights from some of your early product work still guide you today as an investor. speaker-1: Yeah, I think that the biggest thing is empathy with the customer. ⁓ An early example is very influential to me on this. So in one of my first medical device ⁓ technologies that I was working on, my first visit to see a surgeon in the field, in these first cases, ⁓ the device severely malfunctioned. And literally the device went in the middle of an open heart surgery went shooting across the room. Not good, right? You can think of all kinds of surgeons who would not react well to them. It happened consecutively three times in a row in that case. And I think what I really took from this was one, urgency of need for a new solution, two, who the appropriate first customer is for technology, in this case, someone who is very patient. with those early failures. And three, empathy for what the experience has to be in ease of use and how it is the physician is focusing on the patient and what true product requirements are. So I really took that with me through a lot of my experiences in med tech and in technology as well. Empathy. speaker-0: So fast forwarding, you launched Recode Ventures back in 2022. And as I understand it was one of the first dedicated AI driven healthcare VCs. What's the origin story and how did it all come together? speaker-1: Sure. So at the time, I was a partner at my former firm, DCBC, and I sat on a board with another investor from another firm for a long strand DNA synthesis technology company in Cambridge, UK. And we had frequent in-person meetings for the board. And every time I went over to the UK for the board meeting or he came here, we would go looking at companies together. And my partner, Vishal Veladi, and I, who's a physician, came to realize that there was a big opportunity, that ⁓ not enough firms really specialize at the seed stage on healthcare technologies ⁓ in a way that helps companies focus on the patient, helps guide them through regulatory business development, clinical challenges, and so on, and can do that with a focused amount of capital to get them through those early stages and get on their feet. And so Vishal and I left our respective firms and started Recode and made our first investments. speaker-0: And it strikes me that you probably had a new lens on looking at the application of digital and AI to the medtech space, given the empathy and the components of your experiences, really understanding the mindset of the user, that being the physician and then the end user, that being the consumer. So that's probably come together quite nicely for you. speaker-1: Yeah. No, I think that, ⁓ really looking at the details of every step of how your product is can, be used and where it's relieving pain points and where it's causing complication is important. one of the things that I always look for in particularly early stage pre-seed or seed entrepreneurs is the, much they are a student of their market. And so. We have one entrepreneur, for example, that this lane, she was the CEO of probably genetic who knows more about rare diseases than anybody you'll ever encounter. They're patients, the physicians who treat them. ⁓ he has true understanding of the market. And when I see that I say, okay, this is somebody who's special and understands something that the rest of the market may know. speaker-0: So kind of a very deep domain knowledge is an attraction point for you. speaker-1: domain knowledge, but also just how it is that their ⁓ lives, professional lives can be better. And then what it's going to take for someone to buy something to make it better. Right. Because I think often as venture capitalists, we project how we would like the world to be onto a problem that definitely exists, but no one's quite willing to pay for it yet. That's where I've seen failure. So what's really exciting is when I see somebody who has that insight and gets a customer to say I didn't know I needed this Right, but all I know is I'm gonna make purchasing go by it All right. Have you had by the way curious if you've seen that experience too because I know you've been in and out of a lot of surgeries and Some physicians who will say absolutely get this thing out of here and then some of who will say You know, ⁓ my God, this changed mine or my patient's life and that feeling, right? speaker-0: Yeah, and I think there's different types of physicians and you catch them at different stages. So you talked about working with an operator that maybe is too used to working that early innovation. And for those types of operators, what my experience has suggested is that they can see what the future can be, meaning that with certain iterations and improvements that the technology really can get there and be a value add. And a lot of times they're also thinking back to the procedures they've been in. where they've really struggled or faced a challenge that they needed to overcome. And so they're looking for technology that can be inserted in those cases and in those situations. And then from an entrepreneurial perspective, it's a question of, if I add innovation that corrects for this, am I solving for the once in a blue moon or am I solving for the masses? And I think that question is always interesting to explore from what's interesting versus what's investable. speaker-1: Right, absolutely. speaker-0: Yeah. So you're sitting here at the intersection of AI, deep tech, med tech, and healthcare, which means you're just seeing a ton of innovative ideas. Kind of to that point, what companies, what types of companies are you investing in and what separates interesting and just innovation from what's investable? speaker-1: Sure. So I think that the most interesting things I've seen have been where they really understand where the AI is going to be helpful to a customer and where it's just not going to be very necessary. I think that's something that helps something stand out. speaker-0: And so if you were to think about what's in your portfolio now, what are some examples of places where recode has placed a bet? speaker-1: Yeah. So if I were to take a look at something that's really interesting that I find that is more compelling than others, we have an investment in a company called Proprio that allows surgical navigation. So you can predict intraoperatively an outcome of a patient if I position a spine into place A versus place B and what those measurements might look like and what that patient is going to look like coming out of the OR. That is something that we haven't considered doing to date in medical devices. So typically what happens as you know, is that there will be a rep helping to guide the physician with a new technology. They'll kind of estimate it beforehand in preoperative planning. And then during the procedure say, okay, right about there looks right. And the trouble in this case for proprio for spinal fusion patients is 30 % of the time it isn't If you could tell under guidance what that's going to look like post surgery, that's going to be really game changing. But as well, they're also thinking about the post-operative plan where, or post-operatively writing up surgical notes, having training for the physician. speaker-0: Yeah, yeah. Yeah. speaker-1: ⁓ within the procedure, all the things that really get to expanding a market, add additional modules to the technology, and help make the physician's life a little easier so that it keeps them coming back to music. speaker-0: Right. That's a tremendous example. mean, you think about spinal fusion in particular, people often say there's a lot of art that goes in it today and what you're doing is bringing in an additional layer of science. And what I love about the approach is as you gather additional data points, it just makes your intelligence that much better, right? speaker-1: That's right. With each case they do, they're adding to their database. And each data set they get, it makes it more precise and can improve over time. So that's, think, something that's very valuable. speaker-0: Right. And you're capturing both the approach, but then also the outcomes. So you have some longitudinal data, which will suggest, you know, how do we continue to make the algorithm more intelligent and more precise and more predictive, which is really powerful. I'm curious then I've heard some examples in adjacent spaces that would suggest as an operator, if you're approaching a really difficult or challenging, let's say spinal fusion case. You might be able to step into the algorithm and say, you know, how would XYZ expert operator approach this? You know, especially if it's a very specialized case that maybe someone at a very specialized hospital would perform that you could almost say, you know, what's the, what's the approach of the stars that would take sort of to basically take the very high level expertise that one or two physicians in the world might be able to have and basically disseminate that to, to the community is that. an application or approach that you've. speaker-1: That's the strength of the technology. So it's like having thousands of physicians at your disposal and guiding you toward the right placement of devices within the procedure. Today, the way we do that is we have whatever the expertise is of one or two people who are in that OR with you guiding it. But what data has the potential to do is do that for you within the case of the operating room and to provide all those data points of if we place the probe here, will I get to outcome X usually or Y? And rather than relying on Armin standing there in the case with you, right, you can get the consensus opinion of thousands of your colleagues. And that's something that until late, I don't think that we fully grasp as the potential of AI. speaker-0: Right, yeah. And then you think about that potential impact in terms of the ability to potentially democratize medicine and all of a sudden allow for the highest kind of expertise and approach going to everyone versus being reserved to someone who maybe has ⁓ the access to very specialized care. that ability to democratize medicine is a truly powerful concept. speaker-1: One of the potential examples, ⁓ should be clear to say that it's not doing it quite yet, ⁓ thinking of some of the more complex cases like spinal oncology, where physicians today have to ⁓ gently remove tumors from around the spine, moving the spine into different positions in order to access it and so forth. Think of if you could, under proper guidance, understanding if I were to shift the spine into another position, am I going to be able to do that safely? Am I going to get maximum exposure to the tumor? Right now, those procedures are done by a handful of people at some of the top academic medicine communities in the world. My dream is that that can be done at any hospital, community hospitals with the local surgeon, right? With the guidance of data and the potential that that brings. speaker-0: Yeah, yeah, that's an incredible, incredible dream. I'm curious, you hit on something earlier related to payment. I think you said, you we also have to ask the question, who will pay for it? And one of the interesting things with AI applications is that they can be incredibly powerful for the reasons that you described. But then you always come back to how does this become a business model? How does this become something that can actually be commercialized? Is that something that you find yourself asking when evaluating or are you just at such an early stage of investment that you just want to know, okay, can this really improve outcomes? speaker-1: Yeah, so this is a great question and it deserves quite a bit of discussion. So first of all, with anything and let's work with the operating assumption of things that go into an OR, that thinks, you know, for critical procedures. I think it's a little bit different if you're looking at some of these ambulatory surgical centers or ASCs. In those cases, efficiency has equal or more priority toward ⁓ improvement in clinical outcome, just cause they're run a little bit differently. ⁓ however, for most cases in a hospital in the O R improvement in clinical outcome is paramount. I'll stop. Right. And that is your first job to get to. I think that longer term, the entrepreneurs who at the early stage compel me the most are the ones who are thinking about. How do we expand the business model in order to rapidly get out different products in succession ⁓ with our existing technology, including a software solution? And so by that I mean, if I'm going to add a module that's beyond spine and hypothetically think about lung or soft tissue liver, like these sorts of other applications, ⁓ I'm seeing some entrepreneurs ⁓ think about doing this as a software release, which could be approved by the FDA and the FDA has been indeed been supportive of. I think the challenge though is hospitals are accustomed to buying a thing, a piece of hardware. Right. And they often don't know how do I buy this piece of software for a surgical application. They don't think of it in terms of you know, iterations I might get over time, upgrades and so forth and so on. It's more of a physical thing that I buy one time and it's, you know, it's sitting there in inventory. So I think that this is in the rapidly changing environment where we have, agentic AI on the rise and, know, ⁓ thinking about what applications those might bring about. ⁓ I think it's important for entrepreneurs to be including purchasers, hospital executives at the stage where you're developing the product to understand how it is they're going to transact with you because that will influence what you're ultimately building. speaker-0: Yeah, yeah. So the improvement in outcomes is kind of table stakes. It's necessary, but not sufficient. And then you have to move towards, okay, what is the model that potentially could be a line to getting this into an institution, into the hands of physicians and ultimately helping patients. And you're seeing software as an approach that potentially could be the way. speaker-1: Absolutely. speaker-0: That's great. And you mentioned FTA, as kind of device product focused individuals, at least from the outset, we always kind of follow the path of running randomized clinical trials and having that as an approach. then you entered the reimbursement system. But I'm curious from a regulatory perspective, it sounds like you're receiving receptivity to some of these AI approaches that could improve outcomes. speaker-1: Yeah, absolutely. So, and this is going back a few years, but one of my earliest investments in AI needs healthcare and one of the most compelling cases was a company called Caption Health, which eventually got purchased by GE. So let's think about the regulatory journey there. So in order to get an approval, they did four different measures of, and this, for this technology, which was cardiac ultrasound, anybody could do. You could hand somebody a probe. The software would guide them into place like it was ways, giving turn by turn directions, automatically capture an image of the view of the heart you want. And it could be of diagnostic quality. Right? You could train a teenager to do it. ⁓ They wound up getting approvals for medical assistance and above to be able to do this. So think about this, right? A 19 year old with no previous ultrasound experience after a few minutes of training. went out and, as happened in one case, found moderate aortic stenosis in a patient who hadn't been diagnosed with heart disease before. What does that do to our world? So to get an approval, the question was, how is the FDA going to view this? So first of all, I would say that the FDA has been very helpful, actually, and aggressively so. in making sure that there's a pathway for software to get approved. They had four specific clinical dimensions, left atrial size, right atrial size, plural fusion detection, and so forth. think that's the figure out what the fourth one was. But they adjusted for a learning curve. They adjusted for, so they were detecting the AI's ability and not somebody learning ultrasound. They had different types of patients. had, I think, 85 % of them were morbidly obese, something like that. So they worked aggressively early with the FDA to make sure that they could get the appropriate approval, but also a claim that they wanted to have for the Maquette. So indeed, when they got their approval, the FDA issued their own press release about it, which how often you see that? speaker-0: . speaker-1: So normally they're like, ⁓ no, no, this case they were excited to do it. So the good news here is I'd say three things. One, yes, the FDA will work with you. Two, think carefully about your claims that you want to make. Regulatory isn't a thing to get out of the way. It's part of your strategy for what you're going to tell the market because you need that in hand in order to tell the physicians anything. speaker-0: Right. speaker-1: and not only from a ⁓ legal regulatory standpoint, but for your own credibility, right? And there's going to be more studies yet to come after that FDA approval. And then finally, yeah, thinking, working again early with the regulatory agencies about what you have to prove will ultimately impact what you put in that product. speaker-0: Yeah. And so often you have a product you really believe in and you want to get out there and say, it's the best or it's better. But at the end of the day in this vertical, that's not really what wins. It's having appropriately crafted claims that you can really support and prove in a timely fashion that'll get you to your clearance or your approval. I'd love to talk to you a little bit about AI native. It's one of the buzz terminology that we hear a lot. We hear a lot about AI native healthcare. What does AI native mean to you? speaker-1: So I'm going to be a little bit off script and then I'm going to say that I don't, I resist terms like AI native, tech native, tech first, because I think it gets into the common trap that we see with entrepreneurs of tech first, customer later. And that's the thing I used to say, I don't care what's in the product. If it's solving the problem, don't, you don't have to have AI in it to, in order to raise money from me. necessarily like I don't want to, I don't want entrepreneurs to feel that they have to have discussion of AI in order to feel that they're solving a customer problem. What's important is that you're thinking first about how these technologies can help improve those problems. And I think that that's really what's the critical point. So when I think of AI native though, to answer that question, it's from the start, thinking about how I can devise my product so that the data that they are collecting around them all the time is able to help that surgeon or, you know, help inform a patient as the case main being. ⁓ And you are building it in your tech stack from the outset and not at the end as an afterthought, when you have some regressions to tell me about, you know, a potential ⁓ clinical outcome dressing on top. But how did you build that to incorporate it from day one? speaker-0: And then a piece of that probably is building it from the beginning and kind of understanding it prospectively that allows you to collect the data to make improvements as you go along and make it more intelligent. So I'm curious whether or not that aggregation of data is something that you see as a potential moat in a lot of these businesses that get there first or get there early. speaker-1: It's, I'm glad you asked because it's frankly, it's the first thing we looked at. me say we invest in the intersection of AI and healthcare. What I really mean is we invest in companies that have created their own data moat. Let me give you an example from a company I had invested in in the past. There's a company called Epsilera, which had enabled discovery of natural antibodies through a very cool semiconductor like CHIP they had developed. and individual B cells were deposited into these wells. They secreted natural antibodies, millions of them, that can be tweaked accordingly to go after complex human targets. That's a very unique data set because they made the chip. They have the technology to be able to get antibodies from each individual B cell ⁓ that is secreting them. That's a unique way of getting data on millions of antibodies, whereas before we had to give the mouse cancer and watch it supri antibodies. Right. That is unique to them. So I don't know anybody else making that data set. And, you know, as such, they were able to move very rapidly in the course of months, ⁓ had an asset for therapeutic treatment of people with, early stage COVID. ⁓ was one of the two therapies on the market. went from zero to something in the first trials in what, four months, from January to May of 2020. So now some of that, of course, is urgency with the government stepping on it, right? But some of it is just simply how we went that fast ⁓ is borne by the technology. The data set is really the most important. speaker-0: Yeah, that's really, really impressive. And so you're dealing with these founders that are oftentimes very technically strong and you're coming in at an early stage and just for definition. So recodes kind of coming in at a series A or even C. ⁓ speaker-1: Recode is an early stage venture firm. have 10 portfolio companies in our first fund, ranging from about a half million to five million dollars of our first check-in. We do do follow-ons. ⁓ We do lead, we follow, ⁓ and we invest at the pre-seed, seed, and occasionally the series. speaker-0: So you have these technically strong founders you mentioned earlier that you really like people with strong domain expertise and kind of a master of their craft. Do you see oftentimes, especially at the early stages of these companies, mistakes or biases from technically strong founders, especially when they're looking at entering regulated or clinical markets? speaker-1: I think that one of the most common mistakes that is made is that a lot of the entrepreneurs we see will make assumptions about the market they've seen from personal experience in the lab, ⁓ or ⁓ they have a technology where there is a nail looking for a hammer. And the amount of research that they put in is usually on the technical that of striking amazement with the physician. The way I like to see them go about it is talk to three different types of people, your financial, your user, and your technical buyer, because those are often three different people. And speaker-0: Yeah. speaker-1: I think usually the technical founders I see are very good working with the user buyer. They understand that it's fun to work with them. You're getting an improvement in outcome. You're, ⁓ you know, understanding whether this is something that they feel they benefit from and they kind of tell you yes. Right. But left out of that is, you know, the financial buyer, the guy in purchasing or the executives in the hospitals. How is this going to be a priority to them? So there's a difference between something that is, yeah, we would like that, and we absolutely have to have it. Because those customers do not uptake technology at the same rate a bank does in order to defeat the competition. They do it, and these entrepreneurs, think, have to learn that as they fund the product that we work with to say, you know, when They have to learn that hospitals ⁓ operate with a much more constrained budget. They bring things in when they have to. And they're not as engaged with, you technological improvement as we are. And understanding why that is. ⁓ speaker-0: Yeah, it seems as though really understanding the perspective of the business from those three perspectives can be incredibly helpful. And if you can get alignment across all three, buy in from all three, now you have a business model you can run with. what's an example of a business model in AI-driven healthcare that's really impressed you? Or what is the type of business model that you would find to be impressive? speaker-1: So I think things are in a dynamic state of change right now with so much changing in AI so very quickly that I think that the answer to this question would change in a couple months after you broadcast it. So again, stick to the things that you know are going to be important to that person. So for example, If you have something that's going allow that surgeon to get home early and see the kids soccer game, that's a big deal. It doesn't matter if it's AI or something else solving that problem. is a big deal. Right. So always be asking that question early on as to, you know, we've seen advancements in, ⁓ agents that can be used to do tasks for you. Where does that belong? My personal opinion is that ⁓ that will belong somewhere with ⁓ robotics in order to accelerate certain grunt steps of the procedure. That's just something I'm thinking about. But I think that you have to translate that into how is that going to look at improving efficiency in a hospital? How is that going to make buying a little bit easier because they know that they're going to be able to get a certain ROI in this in the hospital? ⁓ And how do, technically, do we know it's going to be safe to have in that a wall? So at each step, I think every time you have something new, you have to say, who does this influence, ⁓ who does this impact, and why is this going to be important to them? I think that's going to be key. speaker-0: Yeah, yeah. And you're getting involved in the early stages, so you're able to shape some of this thinking, but you're also having to think over the long term. So how do you think about structuring capital and supporting differently than maybe a typical VC, especially for these founders with strong scientific perspectives and backgrounds? speaker-1: Yeah, so I think this is very critical to do early on. One of the first things I do with entrepreneurs when they've received financing from us is to say, congratulations, you just got your round closed. Now let's write your next deck for your next pitch. Right? And you get some strange looks and why is that a good investment of my time? And the reason why is because if you can think about those goals for your next fundraising, then all you have to do is wind back and say, here are the things I need to do to hit it. So when we think about things like financing, where are the key milestones that you have to hit? For many, it's how do I just prove that clinically I am superior? And I think that lost upon people sometimes is if you just prove the one thing, just the one thing, forget the rest for now, you win. speaker-0: Yeah. Yes. speaker-1: Right? And the rest you can figure out under financing of public markets or, you know, in a much later stage financing, but trying to do too much too soon and prove that, you know, you can do complex surgeries remotely and, you know, have tons of data and all that by your series A may be a little bit ambitious. speaker-0: Yeah, it's the typical curse of the platform technology. It's yes, share a platform, prove it in one vertical today. And if you win today, then you can see tomorrow. speaker-1: That's right. You get permission to move forward. So let's think about this in device and drug context, because we invest in both. ⁓ On the device side, it's, what are the claims we need to be able to make to the market if we're going to say that our AI-based device is meaningful to them? Who is it meaningful to? And you outline those kinds of claims and then you have to think about to prove that. this is key. How much money do I need to raise? And times I think that when you're thinking of that first clinical stage, people say, okay, what is the, you know, trial I need to do in order to be able to raise just enough money. I don't want to have to raise too much money because people don't want to do that. My investors aren't going to think so that's a red flag. So. You should raise, should do the trial that will prove to the market what you need to prove and not what will make investors happy. And I know that's easier said than done, right? You're an entrepreneur yourself, right? Omar, right? It's like, thanks, you know, I've got to raise this much money to prove, you know, but it really is, you know, if you try to shortchange that, you wind up postponing the problem and have a worse problem later. So I think that you have to look at what are those successive rounds going to look like and what's going to cost to do that actual, that study. so, and thinking about how it is that it needs to be structured and what kinds of people you need to bring on to do it. Then we look at, ⁓ in either in parallel or subsequent to that, what is the data we're going to be collecting from that? And then. How can we use that to make our next claim for our next product? But really butchering those early priorities such that, and you know this from product yourself, right? You have to be ruthless in taking out all the things you'd like to do to just think about the one thing. And tell investors, because if we prove that, we can then go do the next thing. And then over here is what I want. speaker-0: Yeah. Exactly. And I think what I really like about your approach in terms of you close it around, let's work on the next deck. It also gives these very creative personalities the ability to say, look, there's more down the road. Let's focus on the one thing and execute it against that moving the fat and being able to achieve it. But then we can start talking about the other things and we gain the permission to do so. I think it's, it still allows them to tap into that creativity to show that they're. Innovation can be world changing, let's do this in a stepwise prioritized way. speaker-1: Another example that I have is we have a company and I don't want to steal their thunder so I won't say the name, but literally on pre-seed financing from us on less than a million dollars, they are launching next month. it's a ⁓ diagnostic prognostication for ⁓ cancer patients. So leave it at that. speaker-0: Okay, great. You to come back and talk to us more about that. speaker-1: But I'll absolutely do that. But I told them, said, look, this is all I want you to do. I want you to have adoption. That's great that you want it at five centers and have them do repeated cases. for each of them to say, this is how we need to do this then. Right. If you do just that much, you can raise a next route. But, know, and, you know, thankfully they're not doing this, but speaker-0: Yeah. Yeah. speaker-1: you can really get caught up in, but hold wait, you know, we're getting a grant from so-and-so to develop this other product and Dr. So-and-so wants us to build on this and then we can do this over here and while we can get a $30 million customer. Five places. Right. That's all I want to hear about. I said five, right? You know, if they do that and they say, yes, I now know how to treat these particular patients with colorectal cancer. And I know what therapies now to give them. And I'm going to get up at ASCO and say that every doctor should do this. You have a win. speaker-0: I said five. Yeah, yeah. You demonstrated the proverbial fact that the dogs are eating the dog food, same store sales, use whatever analogy or metaphor you want, but this idea that it can be repeated. And then you're also probably learning a lot about workflow and preferences and things that you can then expand. I think that a lot of entrepreneurs, when they get their first win, don't realize how many distractions are going to come in. and how difficult it is to constantly and consistently prioritize and reprioritize such that you don't fall into the distractions. speaker-1: I mean, there can be cases, I think where that happens is where people think that there might be a bigger market that they didn't think about. You know, in that case, I said, okay, keep an open mind, but we quickly, you know, come to think about, is this a priority or not? We, you know, kind of have a little team to say, okay, in a few days, we're going to determine if this has legs to it if we should pursue it and then over. Right. So, speaker-0: right? speaker-1: There can be potential, but I really encourage the discipline first. To be honest, I think 75 % of the time when I see something go wrong with a company, it's around prioritization. speaker-0: That's interesting, really interesting. So what qualities then kind of make an ideal founding team? You mentioned ⁓ dedication to a craft, domain knowledge. When you're looking at a team that you really want to invest in, what are some of the qualities that get you excited? speaker-1: Self-awareness right off the top of my head. And I think that this is a quality that it comes from spending a little bit more time with an entrepreneur than one meeting in my office. It comes from reflection about your experiences. A question that I ask entrepreneurs almost every time at the early stage. And also when I'm interviewing CEOs to come and be a CEO of a company, I've done that before too. is when you look at your last experience, regardless of whether it was successful or not, that's something you might've done differently. It is shocking to me how many people cannot answer that question at all. It really is. I'll get answers around like, well, I should have asked for more money. I should have asserted myself and asking for that promotion. Those are me answers, right? or the company was really successful, there's nothing we should have done differently. Even when it's successful, there's something, right? And if you can't think about that, to say, should have done this a little better, then I don't know what you're going to be like working with to say, okay, what are your shortcomings that we need to surround you with? speaker-0: it. speaker-1: Part of it is I think that we have done, we've done people a disservice and in drumming them up and trying to get everybody to a mythical concept of perfection, right? But part of it is that we need to slow down and say, okay, what are my true strengths and weaknesses? speaker-0: Yeah, well, it's kind of a challenging position for the entrepreneur because to your point, they're almost being trained that they have to be perfect, know every answer, never have any gaps when they're trying to raise funding. And so a lot of them probably think, ⁓ this is a trick question. And they default to the old interview answer of, you know, what could you improve on? It's, well, I could work a little bit less. speaker-1: Right, right, right, yeah. That's always assigned to like, okay, well, you know. speaker-0: To your point that that self-awareness tells you because you know, especially investing at an early stage, these companies are going to have to go through a bunch of pivots and redirection and recourse. And if there's a lack of self-awareness, being able to go through pivots is extraordinarily challenging. So I'm curious, you mentioned before the data mode, which is so powerful. Are you finding that AI is kind of fundamentally disrupting how companies are thinking about intellectual property? speaker-1: Yeah, on this one, I think that it's really in a state of flux. think that it's still yet, ⁓ we have situations where exactly what is patentable has been changing extremely quickly, but it is a question they're asking soon. speaker-0: And with that perspective, because a lot of times people are either coming from a tech perspective or a med tech perspective. For those that are truly med tech founders, are you encouraging them to consider tech investors? It's a big conversation happening right now at med tech is, you know, should we bring in tech or deep tech investors into the space? Do you find yourself kind of bridging the two worlds frequently? speaker-1: So this is a big question. And as you know, I was at a tech firm before I started one and now we are both. Vishal and I come from backgrounds of, gosh, tech, med tech and product. He's a GI physician, gone to venture and to research science. I think that it's really important that Whoever you pick at that person understands what it takes to look, what it takes to get to an outcome where you are. I think that there is, I'd seen some of the challenges that can be on the med tech side are saying, well, you know, development is going very fast and very well. had some great early cases, ⁓ you know, in early trials, and assumptions get made about what the trial is going to look like and, you know, ⁓ what's going to ⁓ yield the positive results and the complexity of recruiting patients and this sort of thing. then when you do get an approval, ⁓ it may or may not mean that people are going to buy it, right? The approval just gets you in the door, right? ⁓ But there's going to be more studies to come. ⁓ Does your investor understand that whether they're a life science or a tech investor, because I think that you need to ask them in order for us to return your fund, what kind of multiple do we need to have? Right. You're going to do a million dollar pre-seed check in my company and you have a billion dollar fund. What do I have to look like for this to look good to you? Right. What is your understanding of how I'm going to have to commercialize? You need to be interviewing them as much as they're interviewing you. People are going to tech investors because their pockets are deeper generally. Right. ⁓ I think that there's, you have the inverse problem. Tech investors are very good at taking business risk. Right. They say a market could look like this. We're going to create a new one and we're going to figure out along the way. exactly how to access it, right? Or what the sales model is going to look like, right? They don't like generally so much ⁓ risk around clinical studies around the product, around science, right? That's a little bit scarier. The life science investors are the opposite. They take all kinds of scientific risk, right? But commercial risk, what do mean? We're just going to sell the product with the sales force and like, that's how it's going to work, right? I think that if we really want to think about making medical devices into very large market technologies, that's going to attract a lot of big dollars. have to think about more than just the current landscape of just building some product that Abbott will buy for $300 million and fit in the sales model. That's what's going to break through. And that's why I think that AI has the potential to remake the medical device industry. speaker-0: It helps them to reimagine market creation in a different way. And if both sides can align in the places that they're strong in, market creation and clinical regulatory risk, then you can really transform a lot of spaces within healthcare. speaker-1: That's right. So let's use the caption health example again, right? Diagnosing anybody anywhere that, you know, an 18 year old medical assistant can do. Well, now think about the market for ultrasound. You know, you can put it in an assisted living facility, nursing home, a mall on, you know, military boats and planes, right? And then just upload that ultrasound and a doctor can, you know, look at it and make a diagnosis. What is... speaker-0: Truly, truly democratizing healthcare. speaker-1: truly democratizing it and ⁓ understanding the elements that we need to make a decision from anywhere. And ⁓ looking at potential payers for that, that go beyond traditional avenues of a hospital, asking the hospital to pay for everything. So thinking about it like an app store, where you can have by the guidance for liver, heart, so on and so forth. get certain quantitative measures around peak flow gradient, right? Ejection fraction, whatever you want. Right. Versus how do I make my in hospital ultrasound machine 10 % better? speaker-0: Right. Yeah. Yeah. And that's where you get to real, really truly disruptive innovation. On a personal note, you recently over the last few years tapped more into your Armenian roots. What has that been like for you personally and professionally? speaker-1: It has been really interesting. So, ⁓ as you know, ⁓ I made my first trip to Armenia last year and, ⁓ it was incredibly eye opening. ⁓ and I went with no real assumptions about what it was going to be like and if you didn't like it. Right. ⁓ and so, ⁓ but an entrepreneur had said, you've got to come speak at the digit tech conference and, by the way, I bought a winery. You should check it out. Okay. So I'll admit the. speaker-0: extra incentive speaker-1: And so, and what I found was pretty amazing that, first of all, a community. The Armenian economy has grown something like 4X in the past eight years. Some of that sadly is due to the Ukraine war and they've had a lot of technical talent come to the country from Russia. But there are real entrepreneurs who they're starting to see some successes with Service Titan, opened up a large office there, Pixar, which is a homegrown company. And they're moving from kind outsourcing economy to a product based technology community, which is really interesting. ⁓ And it reminded me because it felt like when I went there, I had both potential benefits and responsibilities to the community. And I feel that from a personal standpoint, since you asked, think that we need to get more back to that value here, right? Where we realize we are part of a community, people have something to add to it, and that means responsibilities too. speaker-0: Yeah, I love that. And I mean, one of the things that I really enjoy about being part of MedTech is that it's a really strong community. And it's a global community. And the more that you're able to explore and see other countries, it just reminds you that we're all connected in this and we all stand together. that sounds like it's been a tremendous experience for you. And I can imagine something that you'll want to continue to lean into going forward. speaker-1: Absolutely. The new data center that coming from Firebird is building there with 44,000 GPUs, 250 megawatt data center, um, could be transformative to the country in the way that the semiconductor industry was to Taiwan. So, and it's very interesting to see it, uh, totaled out. speaker-0: Yeah. Yeah. Well, as we move to close, maybe we'll turn to the future. And I would just love to ask you, if we were to fast forward 10 years from now and are looking back, what do you think we'll say about what we misunderstood about building enduring healthcare companies in this era? And what do you think we'll say that we got right? speaker-1: Yeah. So this is such a delayed question and not the steel it, sorry. ⁓ but, ⁓ I look at this time with AI and the earlier days as I do internet one.com and around 2000, right? You'll recall in internet one.com dating myself here that, you know, we immediately with the advent of the internet, tried to solve all of our problems, ⁓ with all those companies that would come about. can buy anything online, you know, you know, get your prescription drugs and, ⁓ you know, or from everything from underwear, you know, suits and, everything we tried to do on the internet. Right. Well, what was the crew came out as the real winners, right? The big problem that we had yet to solve was search, right? How do I find the stuff on the internet? ⁓ you know, Amazon focused early on books, ⁓ just books. speaker-0: Yeah, do what? ⁓ speaker-1: focus, right, before moving out from there. So I think that when we think about the age of AI now, I think what's what we look to, what, what will probably people will say about 10 years that we got right and wrong were the, we were right to focus on, you know, thinking about this as a transformative technology because it, were, we were right that it was transformative. This is going to change how we do everything. And that it is not going to be a, a temporary inflection. It is here. So, and if I encourage people with that, if you haven't used quad or chat GPT or, ⁓ you know, vibe coding sites like, you know, lovable perplexity and so forth, I really encourage you to give it a try and see what you can do with it. ⁓ because it, you'll see just how much it can do for you. With that said, I also think people will say that we went too early and try to do trying to do things like automated drug discovery, you know, soup to nuts, that we just needed to nail specific pain point problems. I always tell people that the boring really big problem is a big business, you know, and it should go after that. think thinking about the infrastructure we need. to support AI in organizations is going to be a big deal. So large corporations that want to do their own, use their own data for research and discovery purposes need an infrastructure set up and that's a big business. And I don't think we have enough of those yet. So really focusing on some of the basic needs we have in order to use it, the ability to use it will then follow. speaker-0: Yeah, that's fantastic and it really aligns with how you're thinking about pushing these companies forward from a very early stage. And you know, I would just say on behalf of the community, so excited that you, Recode is leaning in and investing in the pre-seed seed and then Series A to really allow some of these innovations to come to light. speaker-1: It's been my pleasure to be here today and thank you so much for the opportunity and your friendship. speaker-0: Absolutely incredible discussion. And I'd say thank you to all of you for joining us on Powering the Future Within. If this conversation inspired you, follow the show and share it with your MedTech community. I'm Omari Bachnait. Stay inspired and keep innovating. speaker-1: you