Chris, thanks very much for joining us today. It's a real pleasure to have you, and hopefully in the next 45 minutes we will cover some of the topics that are, I think, pretty important, and you've been a leading and driving force behind some of these. So many of the folks in the audience don't know you, so I thought it'd be good if we could just get started by having you describe your journey. It's taken you from an MBBS and an MBA degree to now being the CEO of Viz.ai. What really motivated that journey and really this bridge between medicine and technology? First of all, thank you for having me here today. It's wonderful. We work with many of the innovators in cardiology who are here at the conference. It's great to see some old friends. My name is Chris Manzi. I was a neurosurgeon in the UK, in London, and then my career took a bit of a right turn when I went out to Stanford in 2014 to do the combined MBA and biodesign course. Many of you cardiology innovators will know biodesign because it was started in cardiology by a gentleman called Paul Yock, and so many medical devices have come out of that program. It really focuses on clinical need finding, not just what the cardiologist or the neurosurgeon or the renal physician is doing with the patient that's in front of them, but what happens to get that patient there? What happens after the patient leaves your clinic or your angiosuite or your OR? That really, as a neurosurgeon, opened my eyes to all of the different things that I was missing. I was focused for years and years and years at trying to get good at operating, operating the patient in front of you, and I realized, well, that's fantastic. If I wanted to have an impact more broadly, I needed to take the whole patient journey into consideration. I was at Stanford, Silicon Valley, incredible home of innovation, and it was at a very interesting time. I was there in 2014 to 2016, and it was the age of deep learning. In 2012, something called ImageNet, it's like the Olympics of image recognition or computer vision, had just been won by the first ever deep learning algorithm called AlexNet, and that blew the competitors from Google and Microsoft out of the water. All of a sudden, we could, with degrees of accuracy equivalent to human experts, recognize faces, recognize the difference between a cat and a dog or a dog and a wolf, and we could apply this to medicine in the form of medical images and read medical images accurately. But I realized, with my clinical background, wanting to help the patient, understanding the importance of that journey, if you just applied the technology to an image and you left it at that, you weren't going to make any difference at all to the patient outcome. You might speed up radiologist turnaround time a little bit, like a minute or two. What you needed to do is realize where those patients were coming from, where they needed to get to, which multidisciplinary team members, because medicine is a team sport, were involved in the care of those patients. And therefore, understanding all of that, how could technology make it a little bit better? And so I started playing with the technology, realizing that it could read brain scans, it could read echocardiograms. But then we realized we needed to put it into a system that drove action. And so I pitched the idea in a business school class, and the professor there was Google's Eric Schmidt, who said, oh, this is a really good idea. Hey, here's $2 million. You should go do this. And I thought that was a research grant. Of course, it wasn't. It was venture capital. And so I had to call up my program back in the UK and say, hey, something weird's happened. I think I'm starting a company. And they're like, OK. You're on sabbatical. You've got three years in total. You've taken two. OK. We'll give you an extra year. And then, yeah, come back to the good old NHS after you've had your fun. But we really started to see progress around six months in, where we started to work with clinical innovators in the hospitals that were truly treating these patients. And we realized that this would make a big difference. So we started in ischemic stroke, where just like in STEMI, time really matters. In stroke, every minute is 2 million brain cells. And so we realized if we could reduce the time to treatment, we would improve patient outcomes. And the AI became this amazing trigger that would work the same no matter what time of day, what day of the week, which hospital that patient's in, whether it's we're in Atlanta, which actually was our very first installation, in Grady, which is a public hospital. But it's one of the best stroke hospitals in the world. And it cares for, at the time, 30 or 40 different other hospitals where stroke patients came in and would transfer patients in. We realized if we could get that alert within a minute or so from the scan to the interventionalist at the Grady Hospital, they could drive the workflow and move the patients through faster. I think for me, I've always been obviously in medicine, but I've also been a dabbler with technology. My father was an engineer. My mom used to be a computer scientist and then went to medical school at age 49. So she went the other way. But I realized that if you straddle both those fields, you've got a chance to really drive change. And that's what excites me. I think my favorite people are clinical innovators. And that's why it's great to be here at HRX. Great. And I should have mentioned to the audience that if you haven't done so, please download the HRX 2024 app, and it gives you an opportunity to put in Q&A things that you may be interested in getting from Chris. So Chris, maybe we can take a step back, and many are not going to be familiar with viz.ai. So maybe you could explain a little bit about what viz.ai is and what is its mission in the health care landscape. Yeah. Our mission is to increase access to life-saving treatments. So all of us who work in great institutions, often we have access to colleagues who can help us diagnose, help us manage the patients. We've got the relevant technology we need to treat those patients. But actually, if you look across the population, that's not the case. That's why often for a novel treatment, it takes around 17 years for it to become standard of care, meaning like 80% of patients with that disease are going along that guideline directed care. So we want to solve that problem. We want to reduce the variability in health care. How do we do that? Well, there's the algorithmic part, and then there's the human and workflow part. And even though our name is viz.ai, I would say that the AI part is around 10% to 15% of the value, and the change management and the human workflow part is really where all the value is, the last mile delivery of the insights from the AI. So I'll give you an example. A patient comes in and has an EKG for whatever reason. And viz.ai sits within the DICOM feed, also integrates with the EHR, and will read those images that are coming through, in this case, an EKG. And if we see something that we've got an FDA approval for, like suspected hypertrophic cardiomyopathy or suspected subdural hemorrhage stroke, PFO, then we will send an alert out via application to the relevant team. If it's an acute disease, it will be done acutely, otherwise it'll be added to a list. And that means the specialist team within that health system, not the hospital itself, not just the hospital itself, but the entire health system, now can coordinate the care for that patient. So they set up their protocol for how they want to deal with that patient. Often with HCM, it's going to be get them an echo or check that they've got an echo, get them into a specialist clinic and decide what to do next. But what it does is it means that patients don't fall through the net. It's a safety net, really, population health safety net for the less specialist centers. And it's an accelerator. So you're getting a patient from study to specialist who can then make the actual diagnosis and treatment decision much, much faster than ever before. In stroke, we took it down from five hours to about five minutes. In HCM, it's like five years to five weeks. So it's a different frame, depending on whether it's going to be a chronic disease or an acute disease. But what's happening, and if you look at a lot of our data, for example, the first ever FDA study, that time improvement is very significant. But the reduction in standard deviation or variation is the biggest change. In stroke, we went down from a standard deviation of two hours down to seven minutes. So patients are getting consistent review, consistent review by the specialist to know what to do, and therefore consistent guideline-directed therapy a lot more frequently. This is now, I think, the largest of the clinical AI companies. We started in neuro. We're now in cardiology, oncology, and pulmonology. We have 16 different therapeutic areas or diseases that we work with across 1,700 hospitals. That covers around two-thirds of the US population. And we work with clinicians to try and bring our own algorithms, but also algorithms that they might want to develop onto our platform so more patients can benefit. Whereas for many of us who are working in healthcare systems, we're, of course, inundated on the other side. You know, lots of companies have the promise of being able to use AI on ECGs for disease diagnosis like conditions like hypertrophic cardiomyopathy. What do you fundamentally believe sets your company apart from others in this space? Yeah, I think two things. Number one, the focus on the patient. So it's really easy as clinicians to focus on the doctor, which is very important. But the doctor is part of a team. And as an electrophysiologist or as a nurse, you can't treat a patient without your team around you doing all the things that they do. And so by focusing on the patient, it forces you to take into consideration the nurses, the APPs, the angiotechs, the transfer coordinators, the ED physicians, the ED nurses who move that patient into your world and allow you to make the diagnosis and treatment decision. So number one, focusing on the patient. And we have five values. That's our number one and most important value. And I think number two is we were very early proponents and adopters of implementation science, which means you can't have technological determinism where you throw technology at healthcare and expect it to stick. You have to understand how things work in that hospital, in that health system, whether it's outpatient or inpatient, whether it's a rural setting or a city setting, it varies and it's different. And so we have a large team of former nurses, APPs, coordinators who understand, typically from good centers, but they understand the cutting edge of how the workflow is without the AI, without the technology. And we work very closely with clinicians who want to be at the forefront of innovators to make sure that a technology fits in with the workflow, enhances it, and doesn't get in the way. Because if it does that, we all know doctors will just ignore it and not use it. Well, Chris, a big focus of this year's meeting is implementation science. And I think many of us are struggling with how to acquire and implement these types of technologies into clinical practice. What are some of the larger lessons you've learned, certainly in the cardiac space, about what it takes to implement these things? How do we overcome these barriers to entry? Yeah. So I'll speak from the company side and then from the physician side. So from the company side, we've realized it's very important to understand the technology adoption lifecycle. So you have technology innovators who are trying to solve the problem no matter what, and they see your technology and they steal it out of your hands, and they combine it with other technology, and they do everything they can to solve the problem they want to solve. Then the next stage is the early adopters who see the technology, steal it out of your hands, and use it as it is, and maybe enhance it a bit. And then after that, there's a chasm to the main market, early main market, late main market, and laggards. And many of you would have heard of that work. But if, as a company, you don't focus on those innovators and early adopters early on, you won't get adoption. It won't work. Because when you're first, even for us as a scale company in 1,700 hospitals, whenever we launch a new therapeutic area or a new algorithm, it's new. We don't know exactly where that workflow is still going to break down. Yes, you might be identifying HCM in an ECG, but so what? What happens next? How are the echoes arranged? How are those patients scheduled into clinic? How is genetic testing done? What happens with the family? What is the treatment for these patients? Is there a really aggressive surgical team who really wants to do a myectomy? Is there a team who really believes in cardiac myosin inhibitors? So you've got to understand that framework and work with those innovators to implement, improve your technology. And that takes about a year or so to really do that stage before you jump to the mass market where you're saying, hey, this is ready for broader adoption. And that's why conferences like this, where you can just see the people who want to innovate, want to be in the forefront, and want to work with you in collaboration, why these conferences are so important. From the clinician side, I think if you're one of those innovators who wants new technology, go out and find those companies who are serving your need and help them and work with them. We've done this many times before, but that wasn't the case in 2018, where we were learning how to implement into the DICOM feed, how we were learning how EHR integrations work, how we were learning how IT approvals work. That was so new. And if we didn't have the doctors in Grady or Erlanger, Tennessee or Mount Sinai who adopted really early and said, yeah, that will help you through, we would have failed as a company. And so it's really collaboration from both sides and saying, the world really needs this. This will help our patients. So let's make sure we collaborate to go through the hurdles, to run through walls, and break them down and make sure that we get this adopted. Chris, you mentioned that you started this journey about a decade ago. And obviously, there is a exponential pace with which this is increasing. AI is obviously the buzzword everywhere. How quickly would you say to our colleagues here that you think that AI-based assessment of things will be part of routine clinical practice? Yeah, I think that's one of my main messages, is it is here today. And if in your particular field, there's not great solutions, the technology is there to create those great solutions. You just have to either create them yourself or work with an innovator who wants to create those and can build the technology and get them to you. And the technology works. And it doesn't just work as a nice toy. It works in real clinical practice. We have 1,700 hospitals. I think the vast majority of the hospitals will have another AI solution that they use. And so you don't really have hospitals today who haven't adopted this technology. But within each department, within each subspecialty field, OK, there's a curve for whether these tools are coming on. And so what I would say is, you know the clinical needs. You know what the patient needs. You know the complexities of your specialty, of the workflows and where they're broken. And to find a way to collaborate to ensure that innovation comes to you. In 20 years, well, think about it, 20 years ago, we didn't have any of this, right? And we kind of probably look back at those times as that was the time when we were putting videographs up against the light, at least in the UK we were, right? Now that would be crazy, right? That's the dark ages. Think about what all of us will think about today in 20 years' time. Or because it's improving or accelerating in terms of the innovation, in 10 years' time, we'll be shocked with how limited the technology we had today was. And so if you act with that mindset, adoption of technology is such an obvious thing to do. And the only way to do it is through collaboration. Chris, one of the issues that comes up with integration and implementation is, do you foresee platforms like you embedded into our current EMR systems? Or do you think that they almost become an additional system for managing patients? Because that's been one of the forks that's really been very problematic for people to deal with. Yeah, I think the EHR is phenomenal as a system of record and for billing. And at least the way I practice, you would do your clinical care and then you'd go and type in the notes afterwards, and you might read them before you're doing the care. That's not great for action. And that's why a lot of the alerts that you see reminding you to do guideline-based things in the EHR, I think they're viewed around 8% of the time on average. For VIS, which is not a system of record, it's a system of action, our equivalent AI alerts are viewed 92% of the time and viewed within five minutes. So even with chronic disease, it's something that people are paying attention to and acting on. So I think these two parallel systems have to integrate and have to work well together. I don't believe that the way EHRs are built today, they're set up to be a true driver of action. We are busy in our clinical practice. We go about our day needing to get things done for our patients. And so if the technology doesn't make that easier, we're not going to use it. I'd like to explore some issues that surround, of course, two hot topics, and that's the ethics of AI and also some of the regulatory landscapes. So I wondered if you could give us a little sense about what ethical considerations guide your company in developing and deploying these types of AI-based algorithms in health care, and what are the things that keep you up at night with respect to that? Yeah. So there's lots of parts to that. There's the regulatory part and then the data and ethical part. So great to have the FDA here. In fact, I think one of our lead reviewers, Luke, is here today. So if you're here in the audience, hi, Luke. We work with the FDA a lot. And what we learned very early on, I'll tell a story about our first FDA clearance, because my God, that was hard. Three days before you're going to meet, they send you the feedback from your pre-submission. And the first time we did that, we didn't know what it would look like, but it typically looks the same every time. It's a certain number of pages, six to nine pages of questions, feedback, basically things that they don't like about what you've done with your study or questions they have about your indications for use, et cetera, because remember, they're trying to fit it into a legal framework that they have. But when we were doing this for the first time, those six pages were the most devastating six pages of my life. I was like, the company's over. I've just left my career in neurosurgery. Oh, my God. What am I going to tell my wife? It was awful. And we read it, we cry a little bit, and then we show up to DC. And what had happened for us was the lead reviewer on the medical side had misunderstood how we described the device as being something to help radiologists essentially read scans versus a workflow tool to speed up the entire workflow. So actually, as soon as we'd explained that, all of a sudden they understood the clinical benefit for patients. They understood how much of a benefit this would be. They switched around and they collaborated. And so on the regulatory side, it's really about risk benefit. Making sure that you are benefiting patients. You're not causing harm. On the ethical side, I think the key one is minimizing bias. And I talk about bias quite a lot. And my view is that technology is actually our tool to reduce human bias in health care. And in other worlds, but certainly in health care. Like our health system is biased depending on where that patient shows up. It's not intentional bias 99% of the time. But where that patient shows up, what their socioeconomic status is, who they know, who can maybe connect them to the right doctors. The questions they ask, like how tired the physician is on that day, how specialized that physician is. And so a technology that works the same at 2 in the morning as 2 in the afternoon actually can dramatically reduce that variability, i.e. reducing bias. But you have to build diversity into your training set. And the FDA is actually very good. And I think over the past eight years, we've seen them really adapt their systems to make sure the way they're testing your device actually tests for that bias and tests to ensure diversity. So you've got to have good diversity in your training set and make sure that it reflects the population that you're treating. And so that's different in China than it is in the US. And then you need to make sure that you're focused on patients. Because of the way the economics in different, not just the US, but across the world, are set up, you could have a system that promoted unnecessary spinal surgery, for example. That wouldn't be good for patients. And so you have to have, and you might be able to make money from it, and the hospital might be able to make money from it. You need, as a company, to have water-type values where you're focusing on that patient. And everything you do is around helping that patient and the economics of it come second. So since the FDA is here, what are some of the things you'd like them to be doing to help facilitate the more rapid deployment of these technologies? Yeah, I think when you get to know the FDA, you know that they are there to help the US population become healthier. And that means making sure that they don't clear things that are risky that are going to cause harm, but they do clear things that are going to cause benefit. And so collaborating with them is the most important thing. It's kind of obvious to say, but a lot of people see the FDA as this kind of, not the enemy, but like, oh, I need to be really careful. No, you need to work with them, explain why you're doing it, and they will work with you. One of the things that we've seen over the years is because that's their goal, they're trying to speed up their processes. And actually, it's very efficient. When we submitted our very first clearance, after I cried a little bit and we had a meeting and we finally submitted, we got clearance within four months. And that was via a de novo pathway, which is considered a very burdensome pathway historically. And so the process does work, but speed is important. When you're a company, when you're a clinical innovator, if you have a six to 12 month delay in a clearance, it can mean the difference between your product actually seeing the light of day and not. And so the pre-submission process is this excellent process where you get to get feedback early on before you finalize and submit your study. That's a really good process. If we could, after the first one of those you have, if that could be fast, that'd be fantastic. And then just like they're doing continued innovation with clinicians, they're here to talk to you all to understand where the field's going. And then companies, and they talk to us a lot. So another issue in terms of these algorithms you're developing is, of course, they're based on patient level data. How do you think about the whole issue of patient consent and all of the issues that go into their data being used to define and commercialize these algorithms? Yeah, it's an interesting topic because it's still playing out, not just in health care, but with, say, what OpenAI is doing to create ChatGPT, which many of us will use, an amazing tool. That's been built on data that they certainly didn't pay for. The way it works in health care is early on you collaborate with health systems who, do they own the data? Do the patient own the data? It's actually not very clear today. Certainly the patient has the rights to their own data, has the right to say, don't use my data. But typically, the way it often works is you will be collaborating with Valley, and you'll say, hey, here's a research study, so we're going to use the data for it. Do you agree? You'll get the necessary consents if you need them, depending on the IRB review board. And you'll play through it that way. It is not crystal clear in terms of data ownership how that should work. Chris, I want to shift gears a little bit. You're obviously running a company. Many of us are not running companies, but many may be involved in running departments or things in their organization. I wanted to just kind of get some of your leadership principles. How do you prioritize innovation and maintain a strong organizational culture at viz.ai? So the nice thing in health care is you can have a very strong mission that can attract people to your company who are missionaries, who want to innovate, want to work extremely hard. So that's a benefit that we all have. And then it's about setting those values and the culture around growth mindset, around transparency, meaning any conversation is fine. There's no such thing as a bad idea. Let's talk things out. Let's understand. Let's move forward with pace. In health care, you need to move forward with pace within the constraints of, in a company at least, your quality management system. And within the hospital, within the constraints of the departmental setup and the hospital's rules. So you have to be aware of those frameworks. The thing that I think has always benefited us is we've never worried about where we are today. We've just worried about the gradient of our growth. One of the nicest things we used to say in the early days was, I can't believe how naive we were last week. And it's really nice, actually. We still say that today because that shows that you're really learning about the market. Because as someone running a department or someone running a company, there's so much that we don't know about how things work, both within our own company or out in the community that you're serving or the market that we're trying to sell into. You have to be very humble to understand and learn. Because you're not right. The market's right or the community's right. They have the need out there, and you need to serve it. You mentioned earlier that you look for people who are the early adopters of technology. And I'm sure that, to some extent, you're looking for a similar phenotype for people that you're hiring. What traits do you look for in the ideal employee who's going to push that cycle of innovation and creativity and bridge health care and technology? Yeah, so we have 150 AI-based engineers. We have 80 or so clinical-type people, typically nurses who've worked in hospitals. And they're a very different group, just for two different parts of our organization. But for both, it's kind of the same. You want domain expertise. They need to understand the job they're going to be doing. They need to be experienced at having done something similar before, but have this drive to do more. So we often look for people who, in their career, seem to be progressing a bit faster than the other people who maybe graduated the same time as them. The fact they've come to us show that they're interested in doing something different to what they've done before. We really like clinicians and nurses who come to conferences and speaker stages on particular diseases that are trying to change the field. Because that's what we're trying to do. And so we can actually just become the tool for them to enact change. And then we look for the ability to collaborate. Because the second I left neurosurgery, I'm no longer an expert in what really happens in the cutting edge of that field. Even when I was in it, I certainly wasn't an expert at hardly anything, right? You know what you know. But there's so much more out there in terms of the cutting edge. And so you need a team within Viz that can collaborate with those key opinion leaders who are trying to change the field from within their institutions and are credible enough to be able to create those connections and work together. I'd be remiss if I didn't ask you a question about diversity and inclusion. It's obviously a very important topic for many of us. How do you approach that as the CEO of a company and ensuring that this is addressed even in a technology-laden field? Yeah, very proud of the diversity in our company. Over 50% of our senior leadership team are women, for example. We've got great diversity in terms of ethnicity, skin color, et cetera. And we track that. I think the reason we've done well is we're a merit-based recruiting company. But we always ensure when you're interviewing people for jobs, you have a good diversity of candidates. Then you can choose the very best people who just happen to be diverse in terms of the color of their skin or their socioeconomic status or their sex or gender. And what that allows is a real culture of learning, a culture of diverse opinions that allow you to do things differently to what the status quo might be. Chris, a question from the audience regarding the background of you having worked at the NHS and the US system. The question is, you, of course, know what a very encumbered NHS system looks like, as well as a change-receptive, more entrepreneurial system like we have in the US. How does that impact this whole issue of implementation? Yeah, it's a good question. I think each system around the world has its advantages and disadvantages. I came over to the US, still very proud of the NHS, but thinking it was the world's best system and everything was broken in the US. I've since learned that, actually, yeah, there are things that are broken in the US health system. But there's a reason why it's the leading country in terms of innovation, why you have access to the first devices, the first drugs, et cetera, why you run a lot of the clinical trials. And there's other great countries. Canada, France, Germany, Australia all stand out. On the implementation science front, I think the big benefit you have in the US is you have a focus on excellence versus a focus on average. I think the NHS is very good at trying to increase the average standard across the board. And so there's maybe less variability in the quality of the doctors and the quality of the hospitals. They're a bit more standardized. But we don't see as much, unfortunately, of the sort of power law excellence that you get here. And that excellence in clinicians really drives to that because they want innovation. They want new things. They want to treat their patients more. And so it is actually easier to launch things here than it is elsewhere. Chris, I want to go back. You mentioned hypertrophic cardiomyopathy. It's a disease, obviously, many of us treat and are familiar with. But in the world of AI, I'll just use an example of some of the challenges we're feeling. So obviously, there's no question that we're looking to identify this disease earlier in patients for obvious reasons. But if you go out today, you can purchase an AI-enabled stethoscope that can potentially tell you if someone has HCM. You could use a platform like yours with an ECG, which can tell you if you have HCM. There are companies that will run AI on echo data and will tell you whether a person has HCM. Each of those things, obviously, has different positive predictive values, different points in the cycle. What advice would you have for clinicians if you're thinking of diving in? How would you dive in? Are these all mutually exclusive technologies? You envision them to become complementary technologies? How do you see this playing out? They're complementary. So I would strongly recommend you define the problem you're trying to solve. And you take ownership for solving that. And that might mean using a tool like ours or using, I think you're referring to Echo or us2AI. And it takes a village, both on the technology front, but also within your departments to solve these problems for patients, particularly when it's a disease that was formerly considered rare. 1 in 200 to 300 is not rare. It's just rarely diagnosed and rarely treated. And so we will look back in 10, 20 years and go, how come we're seeing so many more HCM patients? No, they were always there. We just weren't picking them up because they were out in other hospitals or even in our hospital and not with the specialists. And so I would look across the board and put those things together. One thing that we, a big benefit about working with us because of our scale is a lot of other technology vendors, AI companies, work with us and ask us to do the last mile delivery of the workflow change in the hospital. So we have a whole AI Echo platform, which we didn't build. us2AI built it. But that's now deployed through Viz. Big fan, if you're talking about Echo, of that company and what they do with the stethoscope. We're not going to build that. But we'll absolutely collaborate with them. We announced a partnership recently with Cleely. And so we want, if there's technology that a clinician will want and the patient will benefit from, we want to integrate that into our platform and help you access that as a physician. And we'll do the change management or the implementation science to make sure everything sticks together. Chris, you're exposed to lots of cutting edge health care technologies. Are there any that are emerging that particularly excite you at this time? I think the imaging-based technology works. It's fantastic. And it's an amazing trigger. I think some of the large language models, the generative AI, are taking it to the next level. We're incorporating a lot of that into our algorithmic work. And so you will see increasingly imaging-based algorithms combined with algorithms that work on the EHR to get clinical context. So you detect HCM on an EKG? Fine. But you need to know, actually, was that HCM? Or is this patient scheduled for a TAVA procedure next month? Or hopefully not next month. Maybe next week. Is it, like, is this, these technologies come together to really close the gaps that we see in what is often broken workflow? So I think generative AI is very exciting. I think applying AI, like applied AI, is really where we need to focus. The tools are being built by a lot of our technology partners, like Google's big investor in this, and Microsoft, who we work with. And it's exponentially getting better. But the only way to make that work for patients is if clinicians, like everyone in this audience, actually crafts it to serve their patient population. And so applied AI and implementation of AI is what I'm most excited about. Are you at all concerned that hospital systems in general with relatively thin operating margins are grossly under-invested in this technology domain compared to other industries? Hospitals that leverage AI are already outperforming hospitals that don't. They're treating more patients. They're treating those patients more efficiently. They're often getting better outcomes, which might mean lower length of stay. So they're improving margins. But COVID was incredibly challenging. COVID itself, but also the staffing crisis that caused a huge increase in costs. It feels like the hospital systems have gone through that and now are in a more stable, more innovative mindset. And they're starting to adopt these technologies again. And those who do will win. So Chris, in the last few minutes we have, I'd like to just ask you, looking back on your career, what are some of the key lessons you've learned that you'd like to share with the audience? Oh, good question. Be brave. Be resilient. It's hard. Starting a company, if any of you are thinking about doing that, everything we do in our careers actually has got huge challenges. But I think I didn't realize in starting a company just how many challenges there would be, whether it's our first FDA approval or the 13th, whether it's our first CMS reimbursement or scaling a team from 400 to 400. There's so many challenges. So I think resilience and a growth mindset is very important. The thing that I can tell you is true is the thing that, at least for me, the thing that I find really hard and challenging today, in one year's time, I will find routine and easy. And that's happened to me time and time and time again. I'm sure it's the same in EP with different cases, right? Something that was challenging, but you do it, it becomes totally normal. And so just do it, I guess. Just move forward. Make sure you're doing it with the right value system, focusing on your patients. And you will overcome all the challenges. Great. And I guess last, I'd ask you, what do you want to take away from HRX? And what would be your thoughts to the audience about what you think they should take away from HRX this year? Yeah, I think, look, this is such an exciting time to be in medicine. It's such an exciting time to be a medical innovator, no matter what field you're in. The technology is there. And you have this really rare skill set and expertise that really understands what's needed. And so my favorite people in the world, clinical innovators, from this conference, I would personally take that away, conversations with innovators, learning what problems need to be solved, ideas for how we might solve those problems, how we as a company can collaborate, how you guys are thinking about the next generation of companies like mine or device companies. I think the fact that you're here shows that you're a clinical innovator. And I think it's a great time to be there. Well, Chris, thanks very much. I also want to congratulate you. Chris was just named on the Time Magazine's Top 100 AI Innovators in the US. In the US, it's a pretty notable list of who's who, and I think it really speaks to what you're achieving in this field. Thank you for helping us kick off this meeting this year. I think we'll take a break at this point and return in 15 minutes. Just a couple of housekeeping measures. If you haven't done so, please pick up your headsets. You'll need them in the next sessions. And a reminder that we end today with a fun happy hour sponsored by Centiaur at 5.30, and there will be a raffle opportunity at the thing. So enjoy the rest of your conference, and thank you again, Chris.

Chris Manzi

Viz.ai

neurosurgeon

AI in healthcare

ischemic stroke treatment

EHR integration

patient outcomes

ethical AI