Is new hospital technology making clinicians' jobs harder instead of easier? While a wave of digital tools promised to revolutionize care, many have only added to the administrative burden, leading to widespread clinician burnout. This episode dives into a company aiming to reverse that trend with the Artisight smart hospital platform, a system designed to give hospitals digital eyes and ears to automate tasks and let clinicians focus on what matters: patient care. We're joined by Artisight's co-founder and CEO, Dr. Andrew Gostine, to discuss his vision for a "keyboard-free hospital."
We unpack how the Artisight smart hospital platform is transforming hospital operations from the ground up. The platform is an ambient intelligence infrastructure that uses a network of cameras, microphones, and IoT sensors to create a real-time process improvement engine. By combining computer vision, voice recognition, and on-device AI, Artisight automates workflows, enhances patient safety, and boosts efficiency. This approach represents a true clinician-led health tech innovation, born from CEO Dr. Andrew Gostine's own frustrations with clerical burdens during his medical training.
We explore how this technology addresses some of healthcare's most persistent problems. A key focus is AI for patient fall reduction, where the platform has demonstrated the ability to cut fall rates by as much as 90%. Beyond safety, we discuss the critical issue of reducing clinician burnout with technology. By automating tasks like nursing documentation and streamlining communication, Artisight has helped its partner hospitals reduce nurse turnover by over 50%. The conversation also covers the high-stakes environment of the operating room, detailing how operating room optimization AI can improve coordination, automate timestamps, and even create predictive schedules to enhance throughput. Throughout the episode, we analyze Artisight's privacy-first model, which uses on-device processing and temporary databases to ensure HIPAA compliance, and the strategic backing from investors like NVIDIA that is fueling its rapid expansion into hundreds of hospitals.
About Our Guest:
Dr. Andrew Gostine is the co-founder and CEO of Artisight and a practicing critical care anesthesiologist. His journey to founding Artisight was shaped by unique experiences outside of traditional medicine, including running a contract research organization and working in venture capital at a high-frequency trading firm. Witnessing the stark contrast between ultra-low latency technology in finance and the slow, pager-based communication in hospitals sparked his mission. A particularly grueling six-hour ordeal to coordinate patient pre-op checks—a task he later automated with software in just six weeks—was the catalyst for his vision of a smart hospital that could automate clerical and clinical tasks, giving time back to providers.
Learn More From Our Guest / Episode Resources:
Introduction
Dr. Trevor Royce: Paul, I'm dangerously close to you. I've made my way to New England. I'm here at the shores of Cape Cod. Close, but so far.
Dr. Paul Gerrard: About a month too early, I'm going to say. Fall is the better time to be here.
Dr. Trevor Royce: 20 degrees colder than North Carolina.
Dr. Tim Showalter: If you're there in the summer, you get the pleasure of paying through the nose for the place. You don't want to go during shoulder season, because then it's like too nice.
Dr. Trevor Royce: That's true. It's pretty crowded, too.
Dr. Tim Showalter: So you've got the whole crew up there?
Dr Trevor Royce: The whole crew. All the dependents, extended family. I don't sleep.
Dr. Tim Showalter: And you're spending your vacation talking to us on the podcast.
Dr. Trevor Royce: I wouldn't want it any other way. Shall we?
Dr. Tim Showalter: Paul, what's our catchphrase today?
Dr. Paul Gerrard: We're diving deep today. So deep, we needed prior authorization just to record this.
Dr. Tim Showalter: Love it.
Dr. Trevor Royce: Let's kick it. Welcome to HealthTech Remedy, the show where three physician leaders in health technology unpack the stories behind the most transformative health companies and talk with the people shaping the future of care. I'm Trevor Royce, radiation oncologist and researcher with a focus on real-world evidence, AI diagnostics, and informatics.
Dr. Paul Gerrard: I'm Paul Gerrard. I began in physical medicine and rehabilitation and now work at the intersection of reimbursement strategy, molecular diagnostics, and AI-driven market access.
Dr. Tim Showalter: And I'm Tim Showalter, a radiation oncologist and former medical device entrepreneur, now focused on scaling AI for cancer care.
Dr. Trevor Royce: Today, we're diving into a company that may be closer to realizing the vision of a truly, quote, smart hospital than anyone else: Artisight. We'll also speak with its co-founder and CEO, Dr. Andrew Gostine. Super excited about discussing today's company.
Introducing Artisight: The Vision for a Keyboard-Free Smart Hospital
Dr. Paul Gerrard: Artisight is tackling one of the biggest pain points in modern healthcare: the administrative, logistical, and communication burdens that take clinicians away from patient care. Their platform uses ambient AI to automate documentation, monitor workflows, and improve patient safety and efficiency across hospitals. And on an editorial note, make technology help physicians rather than be one more burden on them in the clinical workflow.
Dr. Tim Showalter: They're out there building what they call a keyboard-free hospital vision. They're really not joking. Artisight combines computer vision, voice recognition, and IoT (internet of things) sensors into a unified platform that acts like a real-time process improvement engine.
Dr. Trevor Royce: Yeah, super interesting. This is true health tech, where you're really taking hospitals and just dialing up the technology enablement of it. And they're making it happen. They're live in over 100 hospitals, I think with more than 200 undergoing additional implementation. They do a lot of work on reducing falls. It'll be great to hear from Andrew later with some specific examples. But as I understand it, their AI has reduced patient falls by as much as 90% and reduced nursing overtime by over 50% and improved patient throughput by more than 50% in some cases.
Dr. Paul Gerrard: So what are they exactly? What is Artisight? Let's break it down. They are an ambient intelligence infrastructure for hospitals. Cameras, microphones, and sensors collect data, which is then processed locally using NVIDIA-powered on-device AI to generate real-time insights. The key reason behind this is to make things work seamlessly. We have all experienced new technology coming into healthcare, and a lot of the time it becomes a burden rather than making things smoother because it's different tech systems that don't talk to each other.
Humans have to do that interfacing between the tech systems, and humans are still responsible for all those administrative tasks that are both time-consuming and a lot of us probably just don't like doing. They are trying to use both ambient devices as well as AI to automate these things and make the introduction of technology truly seamless rather than a burden. This includes everything from monitoring if a patient is at risk of falling, which is probably how they're getting those fall reductions, to automatically documenting nursing tasks, to managing OR coordination and clinic flow.
Clinician-Led Innovation and a Privacy-First Approach
Dr. Tim Showalter: You read the description of this company, and I was expecting a pure technologist as the CEO and founder. But what I really love about Artisight is that it's really clinician-led and built by clinicians with a huge amount of empathy for all of the bottlenecks and challenges we face in hospitals. So Andrew Gostine, who we'll speak to later, is the founder and CEO. He's a practicing critical care anesthesiologist who started the company out of frustration with the day-to-day inefficiencies he saw even during his training in the ICU. And I think that runs through the DNA of the company, as we'll find out later.
A good example of this is you start off with this vision of a completely connected smart hospital with all the complexities that involves. But many of their sites that are starting to embrace this technology are doing smaller-scale pilots to have an entry-level embrace of technology and improve specific workflows, all as a step-by-step process to ultimately developing a fully connected hospital.
Dr. Trevor Royce: And I think their platform was obviously built with data protections and privacy in mind. Anytime you're coming this close to certainly patient care, but medical records, you're going to have to be super thoughtful about how you protect data. So as I understand it, no permanent storage of PHI, on-device processing, it's all HIPAA compliant using techniques like temporary databases that wipe each night as part of their data management structure. So it sounds like a thoughtful model that aims to build trust in a very sensitive data environment and respect privacy.
Dr. Tim Showalter: I think this is one of those features that the more I've thought about this privacy structure that they built out, it seems like it checks all the boxes for the stakeholders. You can imagine the IT team, when you say you're going to observe everything in the hospital and it's all going to be fully connected, just how much server space that would require or cloud storage space—it could be ridiculously expensive. Of course, there's probably some liability attached to having all this potential patient data stored. And what if something is missed, for example? They're observing everything and there's some key event that happens. You've now got a record of that happening in your hospital.
If I were a hospital administrator, I'd worry about some of that. They had to build it in this way to really get folks on board in terms of privacy and all the HIPAA risk. But it seems like it also solves a lot of additional problems or concerns in terms of implementation.
Dr. Trevor Royce: Yeah, I think it's a good point. It's a good example of the delicate balance or the delicate dance that health technology plays, where a lot of their tools are going to rely on massive amounts of data. And the more data you have, the better your AI models and so forth can perform, while balancing respect for the individual and patient privacy. This is true across any technology, but obviously particularly true in health technology. So I think you're right, Tim. I think it's a good example of that.
Dr. Paul Gerrard: What stands out to me is how broad their solution set is. We were just talking about things like cameras, which obviously that raises all kinds of privacy concerns absolutely everywhere, especially in hospitals. They have this very broad technology stack. It's not just things like cameras or ambient dictation. It's also things like operating room optimization, asset tracking, and surgical phase detection.
Usually we take the approach, or at least I have the belief that if you want to bring technology to healthcare and improve something in healthcare, start really small and think about how you're going to really improve one narrow thing. I think this is a company that maybe is proving that general philosophy wrong, saying, we're going to look at the whole big picture and improve the big picture. And if you had talked to me a year ago, I would have been skeptical that somebody can do it, but they might actually be doing it.
Dr. Tim Showalter: They're obviously getting noticed. I think there's a lot of traction mounting. You just look at their investment. They had a Series B round on January 11, 2024, which brought in $42 million in investment and was oversubscribed. They have some strategic investors, NVIDIA, and multiple client health systems. So obviously, they're putting together an ecosystem of stakeholders in their cap table, which is pretty clever in terms of implementation and having the support across the complex healthcare landscape.
Host Perspectives: The Wows and Woes of a Full-Stack Hospital Platform
Dr. Trevor Royce: Let's do a quick post check. Give us your wows and woes, Paul.
Dr. Paul Gerrard: For my wow, it's the full-stack infrastructure. They are not just writing software. They're also involved in hardware, sensors, and edge processing tools, all customized for hospital environments and all talking to each other behind the scenes. And for the woe, it's the scope. One, it's a lot to bite off doing all of these different things. The second is they're trying to roll out to many hospitals all at the same time. And implementation at this scale is no small feat. It's not just a matter of what their own engineers are capable of; it's a matter of that interface and interaction with the engineers and the IT staff at the hospital systems they're going to.
Dr. Tim Showalter: When I think about what's really impressive in terms of what wows me about this, top of mind for me is ambient documentation. We're recording this in mid-August 2025. And if you've opened LinkedIn in the past week, all you hear is a bunch of chatter about is Epic going to come in or Cerner going to come in and basically introduce ambient dictation into their own platform and create issues for the Abridge out there, the other companies in that space. There's a lot of attention on that.
When you think about what Artisight can do in that space by having both some capability for listening in to support ambient documentation, but also the camera in the exam room or in the hospital room to document a physical exam or to supplement and enrich the data that's available for the note. I think there's some real opportunity there and it seems to fit in well with their vision of a keyboard-free hospital. I think they've gone upstream or taken a more comprehensive view than I think some of the ambient documentation companies.
I think the area where I would say is a watch-out is integration risk. Getting any solution into a hospital takes a lot of work. So they're going to need to have full buy-in from all of the stakeholders across the continuum. When you're doing something that's multimodal for AI and you need visual data and spoken data and text data, you're going to have to have deep EHR integrations in addition to all these other data types. I understand they have buy-in from NVIDIA and from health systems, but I think one area to watch out for is how they play with the other legacy systems in place.
Dr. Trevor Royce: My wow is kind of similar to yours, Paul. The sheer scope of what they're taking on. Even the phrase "a smart hospital"—that vision is, that could be so many things. That's such a big image in my mind if you think about all the functions a smartphone can do and all the complexities in the hospital as you take that to be a smart hospital. I want to hear from Andrew exactly what that means.
And related to that, the cameras and falls and how you detect falls. I would love to hear from Andrew specifics on what that actually looks like. In my mind, I'm thinking of a Nest camera or a ring camera where you get a little alert because they've noticed an animal or raccoon on your front doorstep or something. Is that kind of what this looks like? Where you've got around-the-clock monitoring in a patient room and you get an alert that's more intelligent than just someone shifting in a bed where you've got some pad underneath and it triggers an alarm. But this is actually a human that's getting up out of the bed, trying to go to the bathroom. Now it's time to activate a nurse. And then downstream of that, how does that impact things like nursing solutions and staffing? So it'll be great to hear from Andrew.
Dr. Paul Gerrard: We'll ask Dr. Gostine about all of this and more coming up, how to build a smart hospital and why ambient intelligence may be the future of care.
Dr. Tim Showalter: Stay tuned for our interview with Dr. Andrew Gostine, founder and CEO of Artisight.
The Origin of Artisight: A Physician's Journey from Frustration to Innovation
Dr. Trevor Royce: Dr. Andrew Gostine, welcome to the show.
Dr. Andrew Gostine: Thanks for having me.
Dr. Trevor Royce: I'll kick things off. We like to start a little bit on the origin story, hear how you came to be what you're doing today. And as I understand it, you're still a practicing physician. Can you tell us a little bit about that and what led you to found and develop Artisight?
Dr. Andrew Gostine: I've always had a foot in and out of medicine. I watched my mom as a nurse and my dad as a physician really enjoy the practice of medicine and went to medical school, I say because I didn't know any better. It was all I ever really knew. But I got to medical school and realized, hey, this was not this perfect, beautiful industry that I thought I was witnessing when my parents were doing it. And so I left and I went to business school and learned a lot about predictive analytics and kind of early editions of artificial intelligence, at least by modern standards.
Then I went back, finished training, and started my internship at Presence Resurrection Medical Center here in Chicago, which was really cool because I had 20 weeks of vacation. It was a really cush transitional year. It was unheard of for interns. And I was bored out of my mind just because 20 weeks of vacation is a lot for an intern, so I started a contract research organization and started validating software and device solutions from startups. I'd bring them into the hospital, run the clinical trials, and got really good at figuring out where the value is.
I continued to go through training and got more and more frustrated with just how it seems we were lagging behind a lot of other industries. But at this point, the CRO company I started was becoming more and more successful. A lot of the companies we were helping were raising a lot of money. And I actually got recruited out of residency to go to venture capital at a high-frequency trading firm here in Chicago called Jump Trading, which is one of the genesis for that Flash Boys: A Wall Street Revolt book. So as part of that, we had a $350 million Evergreen fund, and they tasked us with trying to shave a microsecond off transaction times.
And in doing that, I realized, wow, these people are focused on ultra-low latency communication. When I finished that, I went back to residency and they gave me a pager back. And I realized, these people don't see that there are many, many generations of technology between where high-frequency traders are operating and where hospitals are figuring out how to communicate information with latency times in dozens of minutes. So my first night back on call from this, the vice chair of anesthesia at Northwestern Medicine in downtown Chicago and best hospital in the state of Illinois told me, "Hey, Andrew, since you've been gone, you need to go see all of the inpatients that are in the hospital right now and scheduled to have surgery tomorrow. You need to make sure that there's nothing that's going to delay their surgery."
I said, "OK, how do I do that?" He said, "Go print out the surgery schedule." We do 300 surgeries a day. There's two to three patients per page, about 120 pages. Flip through it, and if you see a number written next to the patient's name, that's their room assignment. I want you to type that all in Excel. Then I want you to go log into another system, see which other residents are on call with you. Then go into another system, get their cell phone numbers, email that out to them, and then call them individually and assign patients to them. The first night, it took us six hours to go see, it was like eight or nine patients. Just the logistics of getting the information took six hours.
I said, "This is crazy. All of this information is in the computer." And the vice chair of anesthesia looked at me and said, "The system's perfect." I said, "The only reason you think it's perfect is because you don't have to do it." The next morning, I called [Could not verify with context]. I'll show you how well this went.
We ended up writing software in six weeks that found all of the patients, found all of their past anesthetic records, made sure all the orders were in, and fully automated all of this in Cerner, so much so that they eliminated the residents from having to do this, and we all got six hours more sleep every night. So it was that moment I realized, holy cow, we could automate a lot of things, if I could just get access to real-time data. And if the data was accurate, I could automate almost every clerical task I'm doing. And someday, a lot of the clinical decision tasks that I'm having to do.
So I started working backwards to myself, where are all the data coming from? And I realized it was from me, seeing something, hearing something, touching a patient, reading their vital signs, reading their EMR. So what if we could make a smart hospital? What if I could give the hospital infrastructure eyes and ears and a sense of touch, the ability to read vital signs? What if we could use large language models to read their medical records and predict what's going to happen? And we ended up doing that, and then COVID hit, and everyone started to realize we need artificial intelligence, remote patient monitoring. And ever since then, it's been this rocket ship and trying to hold on and continue to deploy across now 400 hospitals in the United States.
Dr. Trevor Royce: I love your story there and how, in some ways, it required the exposure to other endeavors like the CRO with idle hands as an intern, and then the high-frequency trading to understand that the hospital didn't have to operate how it always has traditionally. You bringing that outside perspective is that there's a better way, particularly from the data side. We have a lot of young listeners that are still in training and thinking about what to do with their career. And I think these types of experiences really show you how you can make improvements from other industries and bring them into healthcare that can be very ossified traditionally.
Dr. Andrew Gostine: 100%. I think history is littered with tech companies trying to come into health systems and deploy technology. I think the physicians and nurses, really any type of clinical background inside a hospital or even outside the hospitals is so incredibly important. And the people that win in healthcare are the people that understand the workflows to integrate the solutions into our workflows. I think it's just incredibly valuable insight that they have.
Dr. Tim Showalter: I'm picturing that stage of your clinical training was really fascinating because you understood how the hospital worked and you were at the ground level getting the work done. And yet you hadn't been in medicine so long to lose your creativity and just assume that that's how life had to be. Right. That's really fascinating.
Dr. Andrew Gostine: The further you get along in training and practice, the less you are to want to change things. In fact, most people, I think you just get jaded. Most of the solutions that people bring into healthcare make my daily job worse. And so I think you do enough of those and all of a sudden you realize you don't want anything to change because every time it does, it gets worse.
What is a Smart Hospital? From AI Fall Reduction to Multimodal Models
Dr. Tim Showalter: Well, maybe let's spend a little more time on how you re-envisioned care with Artisight. I've read a lot about the smart hospital platform. So when you walk into a hospital that's using some of these technology tools, what would you expect to see? What does that look like?
Dr. Andrew Gostine: It's a story. It's a continuum. And it starts with some very simple things. Just giving the hospital patient rooms and operating rooms cameras so that you can see what's happening. Even when the intelligence that's in that camera is organic human intelligence watching the video feed, there are many, many use cases. Like preventing falls with a camera by having sitters watch cameras. Far more scalable than putting 12 sitters in a room, you can have one person watch 12 cameras.
That solution has been around for a while. You can then take the next iteration of that, which is put an algorithm on that camera that monitors the patients for the staff. You can still have a human in the loop to monitor the video feeds, but with that AI co-pilot, kind of like autonomous driving, it can flag patients where you might need your attention. And now you can monitor 300 patients with a single person. So you're essentially taking that to the extreme of at some point you will not need humans to monitor those video feeds because the algorithms will be far better and far lower cost than paying someone to watch security cameras.
So what Artisight really has perfected is starting with simple use cases and cultivating AI co-pilots on these to scale these models. Another great use case: Cleveland Clinic. They have a camera on that patient and an EEG cap. One tech can monitor four patients because there's a lot of signal to look through. With training a multimodal model to interpret the EEG and video simultaneously, because sometimes if someone shakes a patient, it'll show up as seizure activity, but you then look at the video and you see, no, they're just shaking the patient. So if you have a multimodal model that can interpret both of those simultaneously, you can now scale that to one person monitors 96 patients. And now this shortage of EEG techs that do this monitoring can be allocated to all the patients that really need that resource.
So what you see is us coming into hospitals and these smart hospitals, outfitting them with the sensors, streaming the data for a number of use cases, and then adding co-pilots to automate and scale those remote patient monitoring type hybrid care team models.
Tackling Clinician Burnout and Reducing Nurse Turnover with Technology
Dr. Paul Gerrard: Something that you alluded to just a little bit ago was this notion that healthcare providers in the hospital are so accustomed to new technology coming in and making things more challenging. You were giving me flashbacks of my own time in a hospital. And I remember I worked at a hospital once where they had just gone digital with an electronic health record, but systems didn't talk to each other. So we were printing out information from one computer system to scan it into another. And even within the computer system, things didn't talk. So we had to print out one page from the EMR, scan it into another page. It was like we had technology, but it was actually making our lives more difficult, even though it was supposed to automate things.
And it sounds like Artisight is in fact trying to tackle this problem and do the opposite. But I can imagine that when you're going to hospitals, providers have been traumatized by these kinds of experiences that I'm describing right now. How do you tackle that? How do you make them understand that you truly are bringing in something new that will work with their workflows rather than work against their workflows?
Dr. Andrew Gostine: That's probably the main reason I still practice medicine is so that I am also the end user of my own products at the facilities that I continue to work at. I think a lot of it is just having credibility. So much of this industry is still based on being able to trust the people bringing these technologies in. And my product team will tell you, I harp on them. If they're building a solution that takes 10 clicks when it should take one, they're going to hear from me. And I am the original alpha tester for all of our products before it goes to any client site. I am an easily frustrated clinician. This is probably who becomes entrepreneurs. So if it can't pass me, there's no way I'm going to ever show it to a customer.
First and foremost is just using and taking a dose of your own medicine, as they say, using your own products. Beyond that, it's really helping them to understand the technology. It's a lot of meetings. It's a lot of handholding. It's hiring a lot of clinicians. We have, I think, four physicians and about two dozen nurses on staff at Artisight now. So it's bringing in people that have lived their jobs or continue to live their jobs that can show them how the technology is actually going to make their lives better.
Dr. Trevor Royce: Continuing on that theme, in terms of the end user or the customer, this could be, I guess, patients or hospitals or both. Can you talk a little bit about some of the specific use cases and take that all the way to the outcomes. I was reading about some of the metrics you guys have achieved ahead of time, like I think it was 90% reductions in falls and tens of thousands of nursing hours saved. Can you elaborate a little bit on some of those key use cases and endpoints that you guys have demonstrated?
Dr. Andrew Gostine: The easiest way to understand the types of things we do is to use examples because a smart hospital is really going to transform all workflows. If you think about just your own practice, there are thousands of discrete tasks that you're doing, from documenting to conversations with patients, to ordering tests, to reviewing them, to making a diagnosis. Same thing with nursing. They're putting an IV in one minute, another minute holding a patient's hand, another minute pushing a patient, transporting medication, compounding chemotherapy. There are just thousands of different tasks. And all of them have in common the fact that we need to see something, hear something, collect data, interpret data, and then take an action.
So, a great example of patient rooms and what we do. We'll go in there, we'll put cameras in there. Within two days, we've retrained our algorithms to prevent undetected bed exits. And then we go live. Now we have one person maybe sitting in front of 96 video feeds, but they're only seeing six on their screen, and the algorithms are surfacing video tiles as patients attempt to get out of bed. The problem with hospitals is they'll implement fall prevention measures, and 5% of the patients will get it. And they'll have no falls in those rooms. The 95% of the rooms that don't have someone to monitor them are still incurring falls. So we'll take a hospital and equip essentially a single person to prevent all patients in the hospital from falling, which is why we see when our systems go live, even if they had another video vendor there monitoring to prevent falls, we still see a significant reduction in falls because there are huge populations in that hospital that still don't have monitoring.
Ultimately, the people we're really servicing are the clinicians. We're not really building solutions for the patients. And I would say by far and away, the problem that needs to be solved in healthcare is making the experience better for the doctors and nurses. Ultimately, we're the ones that are doing all of the physical work, all the diagnostic work, the evaluation, the treatment, and delivering that. If all of those people continue to leave, there will be no one delivering healthcare. There is no other industry you can find where they have 30% turnover in their staff.
Artisight, we have less than one person quit a year. If 30% of our team quit every year, that would be catastrophic. That is the norm for hospitals. And so what we've done is really gone in there to make solutions that make the lives better for doctors and nurses. So the stat I'm most proud about is not the patient falls. To be honest with you, I'm frustrated that that's still a problem. It was a problem when my mom was in nursing school. Nobody solved it, and we're still wasting resources on that rather than funding cancer treatments. So what we focus on is making the experience better for doctors and nurses.
We've cut turnover for every client site by over 50%. So instead of losing 25% of your staff every year, you're losing 12%. And that's way more manageable, especially when you onboard a nurse. They go one-to-one with another nurse for 11 weeks. You're paying two nurses to do the job of one nurse while they onboard. That is an incredibly inefficient way to deliver care. That's like ICU-level ratios on the floor. So what we've done is make the experience way better. Now, when you get a team of clinicians that work together, that's how you build like Super Bowl dynasty teams. That's where all the outcomes start to improve that you can't really measure. Patient satisfaction scores start to go up. Pressure injuries start to go down because nurses have more time to do turns. And all of the other subtle things like having time to actually do oral care on an intubated patient to prevent pneumonia. The amount of different outcomes that improve are essentially everything we've ever tried to measure. And it's because you're making the workflows and the experience of providing care significantly better. The derivatives of that are seemingly endless.
Dr. Trevor Royce: The satisfaction of the providers is the other side of the same coin of the outcomes of the patients. To your point, if it's a high-functioning team that's happy, they're working with people they like, they're delivering the high-quality care that they set out to deliver when they went to med school. Then, of course, the patients will ultimately benefit at the end of the day. I think that's absolutely right.
Dr. Andrew Gostine: A hundred percent. It just seems like it was forgotten for too long. And some of the initial ways people tried to solve physician burnout was by moving more tasks to the nurses. And that was not a good solution. It just burned out a different group of people. And so we're really trying to target all the things that burn people out, which is not the lack of a waterfall at a hospital that the doctors and nurses can walk around. It's just the amount of clerical tasks that we have to do that were not why we went to medical school or nursing school. And so it's really unburdening them from the parts of their jobs that they shouldn't be wasting their time on, that artificial intelligence honestly does far better than they will ever do. Taking those tasks off their plate and returning them to the bedside. And we've seen it at every single client site. Outcomes get way better if you allow them to be bedside clinicians.
Dr. Tim Showalter: Well, kudos to you, Andrew, for measuring outcomes for that, because I think the value proposition seems really clear and logical there. If you have happy, engaged providers who are happy with the care that they're providing, of course, it's going to be safer for patients. It's going to be better for patients. And it's great that you're measuring the business outcomes from that.
Dr. Andrew Gostine: Yeah, we just actually published our first peer-reviewed study in the nursing literature out of WellSpan, who's admittedly an investor, but did publish this research before they invested. And that really showed improvements across all outcomes that they measured and delivered a 51% annual IRR, which is a phenomenal financial outcome. I said it before, but I didn't go to medical school because I was really gung-ho about trying to reduce patient falls, but there is a great ROI there. And then once I solve that ROI, I can go after things that might have a more nebulous ROI, like improving the nursing experience. Some CFOs won't buy the ROI on reducing nursing turnover. We don't have to because we can get the ROI from eliminating falls and then I can start providing solutions to them without having to justify the expense of it. So it really is this platform play of putting the right infrastructure in the hospital, getting that initial ROI, and then I get to go solve the problems that I really want to solve, which is making it better for doctors and nurses.
Optimizing the Operating Room with AI: From Coordination to Predictive Scheduling
Dr. Tim Showalter: Andrew, I've spent a fair amount of time in the OR environment as a radiation oncologist who does radiation implant procedures, and I'm sure you've spent a ton of time as an anesthesiologist. I'm curious to learn a little bit more about how Artisight is approaching operating room optimization. It's obviously a very life-or-death type environment, high acuity level. And there's a lot of complexity that happens in the operating room. Can you provide a little bit of context for how you guys are taking that on?
Dr. Andrew Gostine: It's a very similar thesis, not surprisingly, very similar problems in the OR. So we'll go into an operating room, set up cameras that have NVIDIA GPUs inside them to process all of the computer vision and voice recognition there. We'll go in there. And the first thing we do is we just stream the video and audio streams to the OR coordinator desk. You can think of that as the air traffic control tower. And there are people that sit there, and people are adding on surgeries to the schedule every day, and someone will feed a patient, and we have to delay their surgery. And we set up these beautiful schedules for how we're going to do the surgery schedule today, how we're going to do 40 or 50 surgeries. And then, like Tyson says, everyone has a plan until they get punched in the face.
7:30 hits, and we deviate from the plan. And you need something to then re-coordinate those surgeries to send out information to the hundreds, if not thousands, of people that make the ORs run. And so our system will stream video to people who are making the decisions. The algorithms then turn on to track all of the workflows, and we automatically document in Epic the 18 different timestamps that staff have to document.
So when a patient leaves PACU, when a patient enters the OR, we automatically detect that. We send notifications to the surgeon, to the anesthesiologist to get them in the right spot at the right time. Because there are days where I will go to the physician lounge after my first case. They'll enter the OR with my second case, and I won't find out for 20 minutes. I'll just think they're still seeing the patient in pre-op and getting it. And we just wasted 20 minutes. $2,000 just because nobody told me that the patient was in the OR.
We do this real-time coordination, but we also build these really, really accurate data sets of how long was this patient in the OR? How long did it take to do the surgical part of this procedure? We then take that data and we make a prediction engine about how long cases will last. And we build this prediction engine to help you book cases with a far tighter schedule than you would get just using some other EMR tool to book and schedule your cases. So we end up with a schedule that is tighter, more representative of what's actually going to happen that day. And then when the invariable deviation from the plan happens, we re-coordinate to get everyone at the right place at the right time. So much of these hospitals is consolidated and got bigger and bigger and bigger. It's just coordinating to get people in the right place at the right time.
Building a Strategic Ecosystem: The Role of Investors like NVIDIA
Dr. Paul Gerrard: First, I just want to say that quote from Mike Tyson is probably one of my favorite quotes for explaining the realities of healthcare. In a follow-up question, when we look at the information about some of your investors, they include companies like NVIDIA, as well as client health systems that it sounds like have found real value in this. Capital is always good, but these companies sound like they understand the business. We have NVIDIA, who obviously has a lot of expertise on artificial intelligence and technology, health systems who obviously really understand the business of operating a health system. So these are strategic investors. Do you think that having these kinds of investors helps validate your company or give you any additional options for next steps in the future for where you'll go?
Dr. Andrew Gostine: Having been a venture capitalist, I realized early on that venture capitalists can provide a lot more than capital. And the strategic capital that we've taken was very deliberate. In fact, we've only taken small amounts of capital from traditional financial VCs where we felt they could make some other value addition to Artisight besides their capital.
The vast majority of it has come from strategics. So NVIDIA, for the obvious reason that the entire world runs on their chipset right now. But also companies like Johnson Controls, they supply the HVAC systems to 85% of the hospitals in the United States. One of the things you want a smart hospital to do is I want to walk in the room where the patient wants to adjust the temperature in their room. And when I walk in there, I may say the OR is too hot. Please make it two degrees colder. If I can't integrate our voice recognition services with the HVAC system, well, then you're not going to get that kind of efferent arm of that loop to make that change.
So we systematically went after all of the people that build the infrastructure for a smart hospital and universally got all of them to invest, then showed it to the hospitals and got all the health systems to invest so that we could iterate on these things. Because now they have a financial stake in our success. They're going to give us a whole unit where we can experiment on these things. Because if you try to develop something in a lab and then take it to a hospital, I promise you that lab does not accurately reflect what's actually happening in the hospital. So we use live clinical environments for all of our testing and product ideation to make sure that when we go live with something, we know exactly how it's going to work. But that all comes from having a strategic syndicate of investors that will allow us to do that or give us access to their systems to develop the solutions that healthcare really needs.
Dr. Trevor Royce: I'd suspect that has gone a long way, too, in buy-in from the healthcare side. There's so much inertia with healthcare and physicians and hospitals, and it sounds like that approach has resonated. Can you elaborate any more on how you've gotten buy-in?
Dr. Andrew Gostine: You probably see it in your own practice. Healthcare is risk-averse. We don't want to take too much risk because you could have a really bad patient outcome. A physician could lose their license, something like that. You get sued. And I think that's just the culture of medicine is that everyone's afraid of that now. There's also a risk to not changing. And I think more so than ever, that's becoming a real risk. So if all of a sudden you're going to convince a hospital who doesn't have a lot of comfort with artificial intelligence that, "Hey, I'm going to come in there and I'm going to install 3,000 cameras on your network, and then I'm going to train algorithms," it opens up a whole bunch of questions about patient privacy, the network infrastructure, power consumption, heat generation, and climate-controlled environments.
Bandwidth on the networks. There are hundreds of questions that come up. So getting the right early partners where we could show the results in a controlled environment is essentially what it all comes down to. No hospital wants to be first. But if you get a couple of really good partners, people, health systems like Northwestern, Indiana University, Guthrie, WellSpan. All of these systems allow you to take a controlled risk and show the broader healthcare industry, yeah, this works. You can do it safely and it really improves outcomes, financial, patient outcomes, staff outcomes. Then it de-risks it. So getting people aligned that this is really going to be good for them financially, good for their staff, is a way to turbocharge that go-to-market process and the product development process.
Advice for Aspiring Physician Innovators
Dr. Tim Showalter: Well, Andrew, I know we're approaching the end of our scheduled time to talk, and I'd like to point it back towards you and your own career. A lot of our listeners are young physicians or people starting off in their career. Can you give us some advice that you might give to fellow physician innovators who want to build something that impacts healthcare? What have you learned along the way?
Dr. Andrew Gostine: Say yes to every opportunity, for sure. I was the first ever resident, maybe in the United States, I don't know, certainly at Northwestern, to leave residency to go to Venture Capital. I can't even tell you how many of my attendings were like, "You're doing what?" But I took the opportunity and it made all of the difference. Left to go to business school. I think I was the first person in five years at Georgetown to do that. Just follow the path that's less traveled. Get a unique perspective because healthcare needs that more than ever right now. So getting some unique perspective is really what we need to fix it.
Dr. Tim Showalter: Thanks so much, Dr. Andrew Gostine for spending time with us. And that's it for this episode of HealthTech Remedy. Don't forget to subscribe, rate, and share the show. See you next time.
