Healthcare administrative costs consume 25-30% of total spending, yet most AI investments focus on clinical applications rather than operational efficiency. Dr. Yan Chow of Automation Anywhere discusses proven AI use cases delivering measurable ROI today, from revenue cycle management and EOB processing to automated clinical documentation while exploring emerging conversational AI capabilities reshaping patient and provider interactions. As health systems face continued margin pressure, Chow examines where automation investments generate immediate returns versus longer-term strategic value. He addresses the “art of the possible” in administrative AI, implementation realities for enterprise deployments, and why the next wave of healthcare AI may be less about diagnosis and more about eliminating the administrative burden strangling clinical workflows.
Yan Chow, MD, MBAGlobal Healthcare Leader, Automation Anywhere
Megan Antonelli, Chief Executive Officer, HealthIMPACT Live
00:00:00 Intro: Welcome to digital health talks. Each week we meet with healthcare leaders making an immeasurable difference in equity, access and quality. Hear about what tech is worth investing in and what isn't as we focus on the innovations that deliver. Join Megan Antonelli, Jenny Sharp and Shahid Shah for a weekly no BS deep dive on what's really making an impact in healthcare.
00:00:29 Megan Antonelli: Hey everyone! Welcome to Health Impact Lives Digital Health Talks. This is Megan Antonelli and today we are talking about healthcare, AI, administrative automation. And my guest today is Doctor Yan Chow global healthcare leader at Automation Anywhere. He's been a guest on the show before and he is a good friend to Health Impact. Hi doctor Chao, how are you?
00:00:50 Yan Chow: Hi, Megan. It's great to be here again.
00:00:52 Megan Antonelli: Yes, it is so great to have you. And you know, it's amazing to think. I think we really met over, you know, around the pandemic when we started partnering and, and, um, you know, it's been a, a long five years in terms of the progress and the how this pace at which the adoption of automation has happened. So tell us a little bit. I've always, you know, your career and your background and the work you've done. Um, give us a little bit of your background and then, you know, the work that you do at automation anywhere.
00:01:24 Yan Chow: Sure. I'm a pediatrician. I spent most of my career at Kaiser Permanente, uh, for about thirty two years. The last eight years I was the national director of innovation and advanced technology. And from that, you know, leaving Kaiser in twenty fourteen, I went into, um, federal it for a federal contractor as a chief innovation officer for a year or so, and then came back to work for Amgen, a biotech, uh, as the digital medicine lead. And then six years ago, I joined Automation Anywhere technology firm in San Jose, to be there to help launch their healthcare vertical. So it's been, you know, you think about six years ago, it was a very interesting time, you know, right before Covid, right before gen AI, right before all the transparency regulations from federal government. So it's been a wild ride, but I really enjoyed it, thoroughly enjoyed it. And it's always great to see you each year and to find out from other people what's going on. But I think healthcare has been a has been a very interesting journey and will continue to be, I think, the next five or ten years.
00:02:29 Megan Antonelli: Yeah, it's really been so interesting. And I mean, I really think I mean, obviously in our lives and in the world, the pandemic is such a bookend, but in healthcare, it was such a transformational moment. And then that coupled with the, you know, access to, uh, large language models that, you know, really just kind of accelerated everything. And we're seeing that, you know, happen and the adoption happen again. And yet in healthcare, it still takes time. There's still, you know, there's there's fits and starts and there's there's appetites for adoption and then there's real adoption. And I think one of the things about where you sit is you really have a, a very close eye on, on what is and what is not being adopted. Um, and I think we've talked a bit about sort of the revenue cycle and some of the processes that that have been tangible and easy wins for healthcare. But I'd love to hear you talk a little bit about kind of what you're hearing from the hospital and the health system, CFOs and CIOs about, like, the outcomes and the metrics that they are looking for or even what they should be looking for in terms of impact and, and kind of timelines for real value in this.
00:03:41 Yan Chow: Yeah, I think we had discussed before we started, we discuss clinical AI versus administrative AI. And of course, administrators and hospitals tend to look at the administrative AI. I and you think about it. Clinical AI is actually more interesting and it saves lives. But administrative AI saves a healthcare system. So I think right now what we're seeing in the market is that most people are focusing on the administrative AI for now, because that's what they're familiar with in terms of automation. And I would love for AI to do clinical things, but the reality is that US healthcare isn't broken because we lack diagnostic capability. It's collapsing under its operational complexity. So I focus on the administrative side because that's where the need is, where the demand is and what you might call the the cognitive tax. As high as we currently are asking clinicians to be data entry clerks. So when we deploy AI agents to handle prior auth and claims processing or referral management, we aren't just saving money, we're actually buying back mental bandwidth. And so you think about it, uh, automation and AI still is pretty much focused on creating the entire digital body. So the AI is the digital brain that helps us. But the automation piece is what gets things done. And I think that's the interesting thing. So if you're talking about success metrics today and we're talking to CFOs and CIOs, one piece of advice I would give is stop looking at cost reduction and start looking at cash velocity. If a CIO CFO only looks at headcount reduction, they're missing the bigger picture. The metric that matters now more than ever is cash velocity. How fast can we turn a patient encounter into realized revenue? And with the shift to AI agents that handle unstructured data like our document automation module. And these are things like messy jobs or complex denial letters, we are seeing organizations move from managing denials to preventing denials. And so what are the right sort of metrics of success. Uh, one is that you should look at the touchless processing rate. If you are automating. But humans still have to review forty percent of claims you haven't really automated. You just digitized the queue. And so what percentage successfully go through without the need for human review is a metric that is worth looking at. And the second thing is a timeline. With modern gen AI driven automation, the time to value has reduced drastically. Instead of waiting twelve months for an ROI, you should see stabilization in some ROI within twelve to sixteen weeks. It takes longer than probably the scope you have is wrong. And that's what I would say.
00:06:34 Megan Antonelli: Wow. Yeah. I mean, I think that you said so much in that. I mean, I think the, the piece around kind of saving the hospitals. Right. And I think there's and saving the system is so important because without That, you know, we can't, uh, deliver the care that we want to and that we need to. And so and we're seeing hospitals struggle across the country, both in, you know, rural and urban settings for sure. And I think that that, you know, that piece of it is so critical. And then around that, that sort of value and identifying value and, you know, sort of velocity of, of revenue, I think is so important. And when we think about, you know, as you said, the sort of sexy, fun, you know, interesting areas of automating clinical care, um, you know, it's important. But the point you make about what's broken is not our diagnostics. It's not our, you know, research where the we're at the top of the game there. It's the system that is, you know, slow and burdened by so much. And and this back office automation can can help with that, you know quickly. So where do you see you know either now or, you know, looking at the past year or what's to come kind of that those pockets of, you know, innovation around back office automation that is, that are really, um, promising in terms of particularly that, you know, velocity of, of revenue and cash. Cash flow.
00:08:02 Yan Chow: Well, I think, um, traditionally we've seen a lot of automation, you know, in terms of back office and revenue cycle. And we still see the bulk of demand is in revenue cycle, all the subprocesses that are associated with it, you know, such as patient, patient appointments, uh, patient essentially validation coverage validation, the visit, the um, the billing and coding the reimbursement and so on. That process is pretty well worked out, even though it kind of varies from organization to organization. But what's interesting about, uh, sort of the back office automation is it's moving to the front office, and the front office is sort of the beginning of the process where the patient first touches the system. It is also part of revenue cycle, but it's never been a back office. Um, back office, uh, function before. So if you think about it, the traditional back office automation is asynchronous. You know, we can do it at night. We can batch it. But point of care automation has to be synchronous and invisible. So when you actually dealing with a patient that is a critical distinction. So back office automation is working at night. Um front office automation is working in real time. So when a doctor is sitting with a patient, they can't wait for a bot to run a script that's based on information. Even information is available. A few days ago. They need an AI agent next to them or embedded directly in their EHR like epic, Oracle, or Meditech that anticipates their needs in real time using current information, current lab tests, current, uh, you know, patient history in context. So for instance, if you're a doctor and you order an MRI, the agent would instantly check the insurance spot a missing prerequisite, prompt the doctor right then, and prevented denial three weeks later. That's really important. So back office automation fixes problems after they happen. Point of care automation prevents them from entering the system in the first place. And in a sense, I think that's where the market is going. Um, so you're still addressing revenue cycle, but you're being a little bit more proactive right now. Probably the thing that's getting the most attention is auto scribing or ambient clinical intelligence, where you have a microphone in the patient exam room, and then the AI can produce a structured note and recommendations afterwards for the doctor to review. And so we're getting back to the old days, where the doctor pays full attention to the patient without having to consult their PC or do data entry. Now imagine if this experience if in this experience the AI were interactive in real time with the doctor and the patient. The AI would be both a kind of a student intern who can learn from the experience, but also an expert attending who gives advice for the doctor and for the patient. I think you think about the role of AI. I think that's really a very interesting role, and it's just a sort of analogy for the role. It may take in many different areas of healthcare and sort of being an assistant and a learner.
00:11:10 Megan Antonelli: So yeah, and I mean, I think that idea of sort of that automation at the point of care and but you know, where sometimes that might go to completely clinical. What you're talking about is automation at the point of care that allows for that visibility of, you know, coverage, you know, coverage, you know, of capabilities and what what a patient does have access to and what will be cheaper, a less expensive treatment and a higher value treatment for, you know, depending on programs. And I think that that's one of the things that a lot of people, you know, don't think about. But in essence, that's where so much of the friction comes from in a patient experience, right? People don't leave healthcare and say, God, my, you know, my physician was, you know, was terrible. Well, maybe sometimes they do, but generally it's I didn't get coverage or I'm not sure they're giving me what I want, what I, what I should have based on what I can afford. You know, it's this lack of kind of trust so that that idea that AI could provide that visibility and then enable that conversation, which often happens much later, you know, with an administrative person as opposed to with your clinician when it should happen with your clinician. Right.
00:12:24 Yan Chow: And another example of that is sort of a new technology called conversational AI. And this is something that, you know, it's not just, uh, acting with interacting with the patient doctor. It's actually the the interface is the natural language interface that allows a patient doctor to talk in natural language and not have to understand how do you do this so that the machine understands what I'm trying to say? You know, you don't have to do that anymore. So when people think about conversational AI, they think, you know, chatbots. But I'm going to coin a new term, I'll call it what I call the new chatbot is the doo bot. So the doo bot is here with emphasis on the word doo. And so we have all seen conversational AI as they can answer questions like where do I park or what are your hours. But this is low value. The separation between hype and value is agency. So a feature is not a chatbot that tells a patient how to schedule an appointment. It's an AI agent that actually converses with patient to find a time, logs into the EHR, books the slot and triggers the paperwork. And if your AI cannot execute transactions across all the systems that you need to, and it's just a fancy few FAQ page. And the winners will be the systems that connect to connect language models to automation layers that actually do the work. And as much as possible, take these tasks off the plate for clinicians and office workers and patients. So. Right. Yeah. Well we're getting there. We're getting there actually.
00:13:57 Megan Antonelli: Yeah. No, I mean, and that's to go from sort of this AI agent conversation to patient agency, which I think a lot of, a lot of the discussion around patient experience is shifting to patient agency. Right. So that's what patients really want. No one no one expects to go to the hospital and have fun. But they want to be in control, right? They want to be in control. They want to feel that they are making the choices and that they are in the driver's seat. And a lot of the time, it's the lack of information and access to information both on the provider side and the and the patient side. That is in the way of that agency for both of them. And I, um, you know, and even when you think about the physician that that is also their satisfaction criteria is their agency in all of this. Right? If they feel that they're no longer enabled to make the best decisions for their patients because of, you know, sort of black box information into what is or is not covered or things like that. Yeah. You know, um, so it goes both ways. And I think with what you're talking about in terms of that conversational AI, because it's pulling that information and then able to have that discussion with them is huge.
00:15:13 Yan Chow: Theoretically, even ask about conversational AI. Why did you say this? And that is actually a very interesting question, because it gets to the root root issue of explainability. And you don't you're not in control. If you can't explain, you can't figure out why something happened or why you have to do this or that. So I think that's there's a lot of potential as people start designing the systems to really have a two way system that lets you have a lot more visibility and maybe a lot more control into what's going on. Yeah. Yeah.
00:15:46 Megan Antonelli: Well, from your perspective, I mean, there's so much and we've talked about this a little bit in terms of the art of what's possible and the art of the possible in healthcare and AI and automation. And I think, you know, we always get distracted by the shiny objects and what's new and what's hot. Um, but when, you know, you sit in such a unique place in terms of seeing what people are looking at and then a little bit of what they're not looking at, like, you know, sometimes they, they, they're missing. When you think about that, I think when I thought of the question I was putting in eighteen to twenty four month time frame, and that seems too long given the pace at which things are, are kind of appearing. But what's your you know, when you think of what's coming and what's most exciting, where do you think people should be focused?
00:16:33 Yan Chow: Well it's interesting. Um, I would say there are four areas. And we had talked about this a little bit before, four areas I would bet on. And so for the CMO. The CMO Cmio, um, I would say automated clinical decision support at scale, right? Not just alerts, but AI that actively, sort of proactively surfaces relevant information during a workflow. Um, imagine a system that pulls relevant labs and imaging and notes, uh, into a summary when the patient just opens the patient chart. And that would be fantastic. The technology exists, but the challenge is integration and personalization. So health systems that build this capacity will have a significant competitive advantage, not just in quality of care but in recruitment and retention. Because again, you know, for a physician who hears it, some system somewhere has a system where the moment you see a patient, a new patient, it'll give you the entire summary so that you don't walk in ignorant. You know, essentially and missing anything. The second area is for the CMO. And CMO also is autonomous care orchestration. So we're going to I can see a future where we're going to add AI. Everybody is adding AI to their software. So you're going to have a lot of AIS running around. And the question is the issue is how do you know what each one is doing. And this is particularly important because healthcare is a regulated industry. So you can actually build AI agents to act as a um, as for for automated compliance and audit readiness. So you can actually have agents, uh, prepare, um, audits can monitor in real time what's going on with all the other APIs that are running around. And for that, you need actually a governance layer, uh, sort of an orchestration layer. We of course, we provide that, but it's going to be it's going to be even more important because when the regulator says, you know, what is this software doing? You have to be able to explain it and you have to explain why. And that's going to be very, very difficult without a governance layer. And also in terms of security risk and things like that, healthcare systems spend as I mentioned, spend millions of dollars on compliance. And so that's worth doing. The the other area for the CMO and CMO is, um, AI agents can act as patient concierges. So imagine an agent that is activated when a referral is generated and then proactively contacts the patient, finds an opening in the doctor's schedule, handles the auth, and reminds the patient to show up. So this is a win win because it improves clinical adherence and improves patient outcomes. And then finally I mentioned for the CEO and CFO intelligent supply chain and inventory management. This is something you know it's it's amazing. A lot of systems still manual. You know it's still manual and like on Excel spreadsheets. So AI that can predict case volume, optimizes power levels, automates reordering, identifies cost saving substitutions you can actually reduce. Some people say you can reduce supply chain costs by fifteen percent. That's right off the top. So that's pretty important. And it's important to also note that these areas, many of these errors are not sexy. They're not they don't make news, but they do address pain points that senior executives deal with every day. And that organizations I think that invest in these areas will have a months of operational advantage over competitors. So I would recommend in these areas such as, uh, autonomous care orchestration, clinical decision support, intelligent inventory management, supply chain, uh automated uh compliance, and so on that you would invest, uh, about ten to fifteen percent of your automation budget to explore these areas, or partner with vendors willing to do proofs of concept with shared risk. Um, it's interesting. You know, a lot of people think that things would be perfect before you go forward. They don't have to be because there is a learning curve. And so you don't want to be caught behind. You don't you don't want to start when everybody is doing well because at that point you're behind. And so and that learning curve is different for every organization. It's different different resources different strengths and weaknesses and so on.
00:20:53 Megan Antonelli: Right. Well that's a lot. That's a lot of possible. Right. And you know and it's and and sexy or not a lot of it is real. Um, areas where, you know, um, the, the success and failure rate is important, right? If you don't do those things well, you're either losing money or losing patients, you know, significantly. So. But I think one of the things that we hear in terms of, you know, you do these programs, you you know, that there's the implementations tend to, you know, can can stumble a little bit when, um, you know, various things happen. There's a lot of, um, you know, whether you mentioned governance and sort of the change management and, you know, any cultural change and or workflow changes are hard. What are some of the like, you know, if you think of kind of common failure points that you're seeing, what are some of those?
00:21:46 Yan Chow: Yeah, I think the three most common points of failure are, uh, data and technical challenges. You know, because this is a new thing for a lot of organizations, and many of them don't have the expertise. So these will also change with technology changing. Right? So so the points of failure, the data technical challenges will also change. The second thing is a lack of um clinical integration and user trust. And that's part of change management which I'm quite familiar with. And finally the as you mentioned the ethical and regulatory hurdles, the governance, the compliance. So from a process perspective, um, a common issue arises from what some are calling paving the cow path. I like that phrase is taking a broken manual process and automating it exactly as it is. And we see that frequently. What that gives you is a bad process running at the speed of light. And it gives it gives automation a bad name. So successful organizations we've seen do two things differently. First is they do process re-engineering. So they ask with AI based automation, do we even need these multiple steps? Or they may take a step back and re-envision the entire process end to end, focusing on priorities, goals, and results and outcomes rather than accommodating human limitations. And this is this is something that's really pivotal and a lot of organizations I understand. I mean, I understand why they don't have the resources to do this, but long term wise, it's probably the best thing to do. And then the second thing is democratize governance. And so forward thinking organizations, those that are very advanced, uh, don't sort of lock automation in an IT ivory tower. They set up guardrails but not gates. So they allow business units to build their own automations within safety parameters. And the reason is, and I found it's true to be Kaiser where I worked is you wait for central it to build everything. You will never scale fast enough to beat the labor shortage. And guess what? Your staff are already using AI and automation. I bet every organization is doing that and so you just don't know it, so you're not managing it. And I think that's sort of the key. So it it has you know you mentioned AI is a change and AI is a fundamental change. It changes the way we do everything, and it requires the work to really understand how to go forward. And I think a lot of people have their heads down and they're just too busy to do that. Well, it is.
00:24:22 Megan Antonelli: It is, you know, that change and the change management that's required, but it itself is changing so quickly. Right. And I think so we hear a lot of comparisons. I mean the paving the cow path being, you know, sort of the, the direct correlation between, you know, sort of AI implementation and adoption. Right. I mean, that was the last kind of big, huge technological platform change within healthcare. And so we're often creating the comparison between the two. Right. Um, but what are your thoughts on like are they similar. Are they fundamentally different?
00:24:57 Yan Chow: They're fundamentally different. I think AI is actually I would call it the anti EHR. So I think that helps to think of it like that. I think the EHR adoption curve is driven by federal mandates like meaningful use. It was for some clinicians, often without their input, and honestly, it made their lives a lot harder. So it turned doctors into into frustrated typists. And that's and I certainly heard a lot about that. And I felt that to myself when we came up with the EHR. On the other hand, AI adoption is driven by desperation and relief. So AI is the anti anti EHR. Its entire purpose is to remove the friction that the EHR introduced, and is the first technology in twenty years that promises to take work off the clinician's plate rather than adding to it. And because of that, adoption won't be just faster. It'll be bottom up, driven by staff demanding those tools to survive their shifts. You know, it's just an anecdote. When I came out with ChatGPT came out and I was at hims that year. You could not walk twenty feet without hearing somebody talk about ChatGPT. Clinicians and administrators are so excited. So my prediction is that AI adoption will be faster than EHR adoption, but less uniform, primarily because the scope of AI is so much more expansive. So leading health systems will deploy it aggressively across multiple domains and see significant competitive advantage. Mainstream organizations will follow selectively, and I think laggards will find themselves at a serious disadvantage within five to seven years. The key difference is that AI doesn't require replacing core infrastructure because it layers on top. And this lowers barriers to adoption, but also means that organizations can cherry pick applications rather than committing to a wholesale transformation. And whether that's good or bad, it sort of depends on your organizational philosophy, I guess.
00:26:59 Megan Antonelli: Yeah. No, absolutely. And I think your your point around the difference in the distinction, but also that the, the desire to use it to help, you know, they're like you said, people are going to they're going to use it no matter what. So if you don't, you know, if you want to have it, if the governance, then you've got to take, take sort of the reins of, of what it is that you'll be, you know, using um, there. But I also it's almost like, you know, near is a very specific tool and that this can happen anywhere. And in healthcare, people kind of they're always looking for that rule book, that playbook of how do we do this? How do we do this in a, you know, sort of systematized way? What do we start with and what do we end with? And how do we layer seventeen layers of governance over it to, to make sure it happens in a slow and methodical way? Um, but this is much more amorphous. And it comes at every, you know, it's coming from every direction. Um, so it does, uh, does change, change that. And it makes it, you know, certainly exciting to watch and be a part of.
00:27:59 Yan Chow: It changes the way that people do work, which is, you know, it's a fundamental, fundamental change. And so, I mean, I don't know about you, but I'm using AI every day, you know, to to be more productive and actually to do things that I could not do before. I can't spend twenty four hours searching the web for information. AI can do it in a minute, less than a minute. So that kind of productivity multiplication is huge and it just has. It changes the way we think about how we do work. And do we really want to do the work ourselves, or do we want to be able to run several agents to do it for us? It's kind of like a management thing, right? Would you rather manage or would you rather do it yourself? Right.
00:28:41 Megan Antonelli: Well, the orchestration piece is huge. And then, you know, and then it's sort of orchestration and agency a little bit because there is that part of there's a part of the search and research and, you know, sort of finding that you do enjoy. But then is it the synthesis and the outcome that is, is the is the sort of thing that gives you that passion and that joy of what you're creating? I mean, I literally had a conversation with my husband this morning. He's like, you can't even tell the difference between what's I and what's not. And he works in technology and prides himself on all of that. And it's like and I said, well, well then does it matter I don't know. You know, and it's this sort of theoretical thing, like if it's as good as it's as good as you want it to be, then it's good enough.
00:29:22 Yan Chow: You know, it's interesting. A while back, I wrote an article called The Hidden Ripples of AI and the Secondary Consequences of of using AI throughout healthcare. One is that busy physicians will tend to check the box if they know that AI is correct, ninety nine percent of the time, they'll check the box without thinking. And so how do you address that? Because you don't want necessarily one AI to always be, uh, approved. Because what happens is that that goes into the learning database and the AI will learn. That's a correct decision in this case. So it will become more and more towards the medium, right towards the average. And so one of the ideas is to just always present a differential diagnosis of three things. So the physician has to consciously think, oh, I didn't think about this one. Maybe it could be this. Right. And so and there's a lot of stuff that's not in AI that we think is in AI, such as for physicians. We actually when we meet a patient, we sense their posture, their emotion, how they shake hands, you know, how they talk. And that doesn't actually get into AI at this point. Maybe in the future it will, because that's part of what physicians call their gut feel. So, you know, you have this intuition and it's based on years of experience with all these senses and all these sort of outcomes that, you know, so, so it's it's hard to replace that at this point, but maybe eventually it will. And so we'll see.
00:30:56 Megan Antonelli: Yeah. At the American Heart Association meeting in November. On the stage, they did a, um, uh, beat the beat the AI with the physician. So they had three physicians and they had various AI models take questions. Some of the questions were cases that would have been on, say, you know, the medical exams. But then there were some really hard cases and, you know, ChatGPT one. But I think what was interesting, what one, they didn't have ChatGPT and doctors write or, you know, and there were multiple large language models I should favor ChatGPT so much. But they were using all the different models, many different models, plus the physicians alone. And it's the physicians with the AI that we want. You know, that's that's the one that we at least hope to win. Hope wins. Right. But then the other side of that was, you know, when when the AI got it wrong, it got it really wrong. Whereas when the physicians got it wrong, they were and they got it wrong confidently, whereas the physicians got it wrong and they were generally asking for another test. Right? They were saying, well, I'm going to need another MRI. So while the insurance companies might prefer this. Um, it was you know, it was such an amazing thing to watch. And these are like the top cardiologists in the world, you know, up there. Um, and more power to them for actually taking the assignment and doing it with a really good attitude, too. But it is. It'll be so amazing to see how that how that ends up working. But for for sure, we'll be need to be more careful around those decisions than maybe we are with some of the more, um, you know, as we said, less sexy, uh, automation tasks and tools that make a big difference.
00:32:43 Yan Chow: Yeah, that's an ongoing discussion about whether humans alone, humans plus A or AI alone is better. And the latest articles say that it's probably depends on the context, that situation. But, one of the things that is, um, uh, a little bit scary is that if AI is so good at consolidating medical knowledge and so on, um, how can physicians fight back if they have a gut feel that it's different than what the AI says? I think that's, uh, one of the issues that's going to happen is the insurance company may say, you know, why didn't you pick the common the AI recommended route, and now the outcome is bad. And the physician will say, well, because they didn't have to defend, right. They'll just say, that was my my decision based on my experience. Now they'll have to defend. Well, you know, your AI had three hundred thousand cases to look at. I looked at three thousand in my lifetime. So I think I think it's really some of these issues are. And then the other issue is kind of interesting in terms of the AI being expert, is that if you think about it, if the AI can solve issues like sore throats, urinary infections without your help. Then what is the physician left with? The physicians? Whole schedule is going to be cases that cannot be solved by AI. All right. So that's a very daunting kind of schedule.
00:34:06 Megan Antonelli: They're going to be wishing they had their EHR in their pajama time back.
00:34:10 Yan Chow: They have no break right. They have no break. Every patient is a complex case right. And so you'll have cognitive burnout. You know, you have a different kind of burnout. Right. And so I, I think a lot of these issues need to be discussed. And, um, you know, it need to be solved. Otherwise humans, uh, there has a there's a human component when you redesign systems and that that implies you've thought about what is the value of a doctor of a nurse. You know, what's the value of this and that? Uh, and where can we use those people? The best. And I think, you know, I don't think people are talking about laying people off, but I think it's just a matter of giving all these new technologies. Uh, what's the best way that still meets their needs, their needs for professional satisfaction and and, um, a survivable career. So yeah. Yeah.
00:35:01 Megan Antonelli: No, absolutely. There's there's so much there. And I think, you know, and that's where, where I worry is making sure that we continue to have the ability for that deep thought because these are complicated sort of decisions that that we can't outsource to the algorithm that we have to put, you know, value based decisions on in how we design these tools to help and what we want to keep for ourselves versus, uh, automate away. So it's so interesting. A great place to talk about that will be health impact in February.
00:35:32 Yan Chow: Yes, yes. Looking forward to it.
00:35:34 Megan Antonelli: You know how we in our last few minutes, we always like to talk about what's good. And we've talked about lots of things that are good. But if you just have to pick one thing, what are you most kind of optimistic about right now as we look into AI and and automation?
00:35:49 Yan Chow: And I think, yeah, that's actually the picture looks good. Actually, it's becoming good. And first is that AI is definitely having an impact on clinician burnout. I think people are very happy with gen AI and stuff. I think second, it's clear that the economics are undeniable. There's a lot of benefit, clear ROI to automation and AI. And then the third thing, which is even more intangible, is that the technology is advancing faster than expected. And I think that's why it's hard to predict, hard to plan for the future. What basically is the problem though is implementation. So healthcare, we just don't have expertise very much in deploying automation and AI. It's a new thing right? So so as a result you know we at automation, we actually are developing new programs for advisory and consulting services because we're finding that people are asking. And, um, you know, we're one of the resources, maybe for knowledge, but there's very little knowledge out there even among, you know, big organizations. So ultimately, I think my optimism isn't about technology, it's about necessity. And that that actually is a very heavy driver. So and, um, but the reward is worth it. I think, you know, we're working towards the future where clinicians get their time back. Organizations can thrive without downsizing and patients get the access they need. And that starts today. We have to take the first steps today.
00:37:19 Megan Antonelli: Oh that's.
00:37:20 Yan Chow: Great. If the audience is watching this podcast, we would of course love to chat with you and, um, and get the process started. So yeah, you have to do.
00:37:27 Megan Antonelli: Well, yeah. And I was just going to say, you know, I always learn so much when I have a chat with you, um, either on, you know, on our, our platform here or in person. And so, uh, the fact that you guys are going into sort of that consulting and advisory services is amazing because I think you guys just have a wealth of experience that, Um, you know, it's hard to get, you know, sort of that real, real world view of how to implement these in practice. So, um, that's great. And I do, um, welcome and encourage our, uh, audience to both tune in and come to health impact. Um, but to reach out to Doctor Chao because he is, um, a fantastic resource in this space. So thanks so much for joining us. Um, and I look forward to seeing you soon. Um, and to our listeners, you know, please visit Automation Anywhere. Com to learn more about their work. Um, and always reach out. And thanks for more conversations with healthcare leaders. Tune in, subscribe to Digital Health Talks and visit us at Health Impact Live. This is Megan Antonelli signing off.
00:38:29 Outro: Thank you for joining us on Digital Health Talks, where we explore the intersection of healthcare and technology with leaders who are transforming patient care. This episode was brought to you by our valued program partners. Automation Anywhere, revolutionizing healthcare workflows through intelligent automation. Natera advancing contactless vital signs. Monitoring elite groups, delivering strategic healthcare IT solutions. Sailpoint securing healthcare identity management and access governance. Your engagement helps drive the future of healthcare innovation. Subscribe to Digital Health Talks on your preferred podcast platform. Share these insights with your network and follow us on LinkedIn for exclusive content and updates. Ready to connect with healthcare technology leaders in person? Join us at the next Health Impact event. Visit Health Impact forums for dates and registration. Until next time. This is digital health talks where changemakers come together to fix healthcare.