AI in Healthcare: Why Humans Still Matter with Brandon Fertig, Senior Manager at Philips | Ep. 62

In this episode of the MedDevice Cyber Podcast, hosts Trevor Slattery and Christian Espinosa are joined by Brandon Fertig, Senior Manager at Philips, to discuss the intersection of military experience, healthcare cybersecurity, and the evolving role of artificial intelligence. Brandon shares his unique career trajectory, which began with a 20-year tenure in the Air Force where he started in IT during the early days of networking and cybersecurity before transitioning to flying C-130s. This dual background in hands-on IT and high-stakes aviation provided him with a deep appreciation for discipline, process, and the critical importance of addressing the "human factor," skills he now applies to the medical technology (MedTech) industry at Philips and through his work with MedTech Vets, an organization that helps veterans transition into the field. The conversation delves into the core challenges facing the MedTech industry, particularly the tendency to treat cybersecurity as a reactive measure rather than a proactive, integral part of the product design lifecycle. The hosts and guest argue that robust cybersecurity is not merely a technical requirement but is fundamental evidence of quality code and is directly linked to patient safety. They explore the concept of the "Swiss cheese model" of risk, where multiple seemingly small oversights in planning—from regulatory to security—can align to create catastrophic failures. This is exacerbated by the human element; project timelines are often drastically underestimated because the complexities and edge cases are not accounted for early in the process. The speakers emphasize that what may seem like a one-month project can easily extend to six months or more when security, regulatory, and cross-departmental coordination are belatedly introduced, leading to delays, increased costs, and frustrated customers. A significant portion of the discussion is dedicated to the impact of AI on the modern workforce. The panel counters the common fear of AI causing job displacement, reframing it as a powerful tool for augmentation and efficiency. They argue that professionals who learn to leverage AI to automate repeatable tasks, analyze data, and streamline processes will become more valuable and irreplaceable in their roles. However, they also stress AI's limitations, particularly in handling novel or un-trained scenarios, where human creativity, critical thinking, and nuance are indispensable. Using the example of a Waymo self-driving car getting stuck in an unusual traffic situation, they illustrate that while AI is excellent at procedural tasks, the human ability to connect disparate dots and solve unique problems remains a vital checkpoint, especially in critical fields like healthcare and aviation.

Christian: people are afraid of AI taking their job, but if you can learn to use AI to your benefit, learn to prompt it properly, set up your custom GPT, whatever you need to do, it will help you add more value to whoever you're trying to add the value to as well. Brandon: Grab a hold of AI because it ain't going nowhere. It's only going to get bigger. Brandon: I want to get so good at AI that I'm irreplaceable. Trevor: And if you're putting quality training in, you're going to get quality output out of it. If you don't have any training on a situation, you're probably not going to get any output from it, and your Waymo is going to just go back and forth in the middle of the street. Brandon: I think it comes back to something where you can never remove it. It's the human factors. Brandon: leaning so heavily into AI that we start missing the whole point of it in the first place is to build in patient safety and quality. Trevor: Cyber security is evidence of quality code. Full stop. Trevor: Hello and welcome back to the Med Device Cyber podcast. Today, we're going to take you down a journey to see where the intersection of healthcare cyber security and some military experience can come in to tie everything together. I'm joined here by our co-host Christian Espinosa and we have a very special guest, Brandon Fertig. Brandon, I'll turn it over to you to introduce yourself here. Brandon: Hey, thanks Trevor. I'm really glad to have made this connection. Christian and I met at the MedTech World, so I want to give them a plug for their ability to connect people of like-minded backgrounds and, um, and create and generate excitement around a lot of the work that we do. Christian and I kind of talked at some, uh, Mettech conference in San Diego. and I know we've talked, uh, online about the Mettech World Conference and, uh, unfortunately for me that remained remote due to the government shutdown and some of the nuances of what happened last year. So I was trying to make it out to the world conference but I'm going to try to do that again very soon, uh, Christian. So, you know, as you guys asked me to be here today, uh, just my background is I did, uh, 20 years in the Air Force. I started off in IT. So, uh, Christian and I share, I think, a appreciation for the military when we kind of started to cut our teeth. I mean, our networks were, were in their infam, their infamacy, infants, infancy. And, um, you know, they were, they were just coming out of like a, you know, a lot of things were still text-based, especially in the military. ADA, you know, we used ADA for programming or a lot of govern- programming, things of that nature. And they were, you know, the forefront of setting up domains, DMZs, firewalls. Security was a huge thing that everybody was projecting was going to be a huge job for everybody, and it was, and it is today. So, cutting my teeth on that really gave me an opportunity to see a broader scope of what happens behind the scenes, as you get into leadership, as you get into management. You have a great perspective of what it takes to get projects running, and security is always a major factor. So, uh, when Christian and I, um, discussed this podcast, I think we both found that we're going to have a lot to talk about. I'm very excited to be here, so thanks for having me. Christian: Well, thanks for joining. Where are you coming from today, Brandon? Brandon: Yes, I work remotely, so, um, I do like to travel. I've been to 60 countries for the Air Force, but also for my own enjoyment. I travel extensively. But today I’m coming from Arkansas. I set up camp here mainly because I used to teach here at the C-130 schoolhouse. I used to fly C-130s in the Air Force. And, uh, coming here, it was based on a network need, but now that I've been here, we've made huge leaps and bounds in the medical space here. And that actually was a secondary effect of me just being in a project management role and also being central to the whole country. Arkansas is an affordable place to live, has a lot going for it. It's big growth, big vision here. And also, I think people are looking to find better ways to do business and save money. And Arkansas is been really hot for that right now. A lot of business moving in here, technology, AI businesses, security and healthcare are all huge right now in the, in the metro Little Rock area. And, uh, just happened to be in the right place at the right time. So I come to you from, uh, the Midwest of, kind of considered South Midwest of America. It allows me to get on a plane and jump anywhere I need to be in a moment's notice, and it's, it's really flexible. So we use a first springboard and, uh, we love living here. Christian: Yeah, awesome. I graduated high school in Arkansas. I moved to Arkansas when I was 12 from California. Small world. And then I moved to a small town in northwest Arkansas called Clarksville. Well, it was actually called Knoxville. It had like 800 people in it. And I developed a thick southern accent. It took me a year to get rid of that accent when I went to college. Uh, people made fun of my accent at the Air Force Academy quite a bit, but I, I got rid of it. I think I got rid of it. Every now and then it shows up, especially if I go back to Arkansas. Brandon: Yeah, I tend to, I tend to kind of get around my audience. I try not to, it's not like you're making fun of the accent, but you kind of get plugged into what people are talking like. And I'm from the Northeast, I'm originally from Pittsburgh. So a lot of the yinzers will come out every now and then about, you know, a crick or a washcloth is a worshcloth. It’s just little things like roof and roof. You know, America's a great like that. Like you can go anywhere and you'll you'll hear a little bit of something where somebody's from. But, uh, the Air Force kind of got rid of that for me. Christian: Yeah, the Air Force got rid of it for me too. My, uh, wife's mother is from Arkansas and she says 'worsher'. 'The dishworsher', you know? Brandon: Yeah, that's her. That is the proper way. Christian: Right. Well, cool. And thanks for your service. I know you flew for a while as well as did, uh, cybersecurity in, in the Air Force. Is that right? Brandon: Yeah, I had a really crazy beginning. It, it really started off being in, uh, my brother was in computers. He's a savant. He's one of the smartest people I've ever met, and he kind of opened the door for me when he got out of high school to be like, hey, you, you know, this is what I'm getting into and it's, it's huge. So it was, again, somebody leading and I was willing to follow, and it started with my brother. So I grew up kind of very humble means, kind of the American story, you know, low-income housing and had a single mom. But I didn't have a lot of opportunity. The Air Force gave me that and had a brother who was in IT. So those two came together and then I needed to, to have a way to move forward and to make something of my life. And IT gave me everything, it really did. So, starting off, it was really eye opening as to what you can get into if you have the time and the desire to put that work into it. And I think it, anybody can do IT, but I think some people do have a knack for it. It came in the family. So my brother kind of opened that door for me and then I, I started pursuing college and then college led to 9/11. 9/11 I joined the military straight after that. I was just felt like I needed to go. And so, in the first four years I built out domains in Charleston, South Carolina. Uh, I was then in Korea for a year and built out geographically separated units that came with VPN work, that came with, uh, getting knowledge on, you know, virtual private networks and and building LANs and WANs and how those affected you and security wise, especially in Korea. And this is early 2000s, so a lot of stuff was just uncoordinated. There was no quality management system behind it. So a lot of the things we made up were, you know, it was grassroots comm guys coming in there and be like, hey, I'm just a computer guy, but this is what I think we should do. And the Air Force was like, okay. So nobody was, nobody anywhere was, was really all that well coordinated in the beginning years. And then, uh, you know, eventually we, we centralized to Scott Air Force base out of Illinois and Scott Air Force base, you know, they started to get smart on this and they were like, we need to, we need to have every domain look the same so that everything was, you know, understood and you could teach your workforce to work around that, which is ironic. I work in teaching a workforce now for MedTech. But, you know, those early years we were making it up as we, we were going, Trevor and Christian, you know, because we didn't have a set of standards, which I thought, you know, when, when you're young, you're like, I don't really feel like I'm bringing a lot of value here, but in hindsight that was my formative years. It was like working in the unknown. So I started off that way and then aviation kind of came in my knowledge about, you know, you can fly in the Air Force. I wasn't an officer, so I didn't think that was, I didn't want to go back to school right away for that. I was more IT driven, more management driven. So I started flying in the Air Force as an enlisted member. And, uh, when I started doing that, it took me all over the world. It taught me how to, you know, be dynamic and to learn. Went on to get my bachelor's and master's. And I think the aircrew world gave me confidence that I didn't maybe didn't, what, wasn't going to get in IT at an office. So it really kind of developed me over time. It was like, okay, I got the background, I got this management style, and then it grew me to who I was. And then when I got out of the military, you know, Philips picked me right up. They were like, yeah, you have all these skills we need. Project management is where you're at. So again, I followed what people were telling me and listened very carefully to what people were telling me. Christian: So the military skills helped you get into Mettech? Brandon: It did. It did, it paved the whole way. Christian: Cool. And uh you're also part of MedTech vets uh as well, is that right? Brandon: Yeah, and the MedTech vets was really my way to give back. So as I transitioned out of the military, kind of give you maybe some more, more of that that story, when I came out, I knew I would, you know, for me it was kind of a really weird time. We, I came in right after 9/11 and we left, we left Afghanistan in 2021, which is almost a 20-year anniversary. And that was the 20 years I served. So, you know, in the military, you can you can go, you can take a full retirement at 20 years. So, I knew how much time I had had left and about two years before I separated, I started to really key in on what was important to me. I had a master's degree. In fact, only .8% of the Air Force at the time I looked that statistic up. I think they did a a census in like 2017 that only .18% or .08% of the Air Force actually has a masters. So I was like, well, man, I'm bringing a lot of value. I think there's something else I need to be looking at. And then that confidence started to come and then I got my PMP. And then Philips was kind of that transition period where it was like, I don't know what I'm going to do, but I had a brother who worked at Philips. Again, the same guy who kind of gave me that confidence of that older brother uh said, hey, what you come look over here? And you know, I wasn't medical. I think that's one thing we need to tell our audience is that you don't necessarily have to be medical to work in IT and in MedTech because the transition for you is can you lead a team? Can you be a problem solver? That's the huge one. Can you talk the talk? Can you understand complex problems and kind of understand what your customers are going through? And if you can make all those things happen and then get in an interview where you explain what you would do, people are going to see your value right away. So I got picked up within three weeks of retirement. So I had a retirement and then I was like, okay, let's go. And then Philips hired me in and gave me my first shot. I've given them the last four years because of that, there's discipline that comes at Philips with when you get in there, you appreciate, especially from a military perspective, that someone gave you your first shot. And I gave them everything, and they've taken good care of me since then. So there's been a promotion in there, there's been a lot of exciting work, and then I just generally on the outside, I try to bring that same value to MedTech vets, to give back to veterans, to give back to others, raise money for veterans by climbing Kilimanjaro last year, and just doing what I just know that I enjoy and just being myself. And those things all came together at Philips. Christian: That's awesome. Uh, Trevor just uh, made uh, was promoted, I don't know when it was, Trevor to COO. So he's Trevor: Couple months back. Christian: Yeah, so he's uh, in the hot seat with leadership and everything else. So, uh, what do you, what do you think, Trevor, like, a lot of what Brandon was saying about all the skills you need to learn as a leader and all that to bring to the table? What have, with the transition to COO, what have you found? Trevor: I found that uh, I probably could have used some of those skills and the discipline picked up in the Air Force. I did not go down that path, so it feels like at some points flying a little bit blind, but I think that, you know, going up through a process, going up through a company different at the enterprise level and at the startup level, but being able to see your own growth and control that, I think is a really powerful thing. And even expanding past from, you know, before being hands-on keyboard, executing most of the tests that we were doing up to this point is giving me a bit of a perspective where I understand where the process came from initially. I've seen it, I've been part of it for so long that I think it's giving me a bit of a good insight where I'm trying to translate those skills. And uh the rest of the skills becoming an effective COO, that's going to be a work in progress, but uh trying to get a little better at it every day. Christian: Yeah, in our company, I don't know how many we got, uh maybe, I don't know, like five or six veterans, I think. Uh one from the British army, so he's a little bit different. but yeah, I think veterans bring a certain unique discipline to the workforce. uh and uh they're used to working under pressure and then most of the time they stay fairly calm. I don't know, what, what do you think about it? Is that true Trevor, based on the veterans we have in our company? Trevor: I think so. And, you know, for, for an example, our director of project management is a veteran and she's able to proceduralize and take any of the questions or any of the gaps out of a process to an incredible effectiveness. And so if there's ever any question, how do we do this? You know, where's the wiggle room there? Those questions go away. And that is such an important skill. I think especially within the project management realm and it's really important, it's really effective to see. I don't think that it's an easy skill to have either, understanding how to poke the holes in a process is it's not going to be an intuitive thing. Just following a process and saying, 'oh well, you know, it's not perfect but what is' is so much easier then saying what can we do to make it perfect? Or at least, you know, 95% of the way there. Christian: Yeah, solving the the problems and coming up with a standard operating procedure and removing the ambiguity is definitely a skill. And that's something, you know, you have to have in the military if you want to have a bunch of people on the same mission and line, they have to know what the operating procedures are and everything else. Trevor: It makes me think back to what you were saying earlier, Brandon about kind of the early days of, like the early 2000s with IT when it was a little bit of the Wild West. There was no standard process for going through things. And I wonder, you know, obviously it's not as much the case anymore. There's been cybersecurity has been a bit more of a battle tested industry. It still changes really fast, but I'd be curious to hear your thoughts on how the more proceduralization and I guess formalizing cybersecurity and IT in general has been beneficial both for the good guys and for the bad guys. I mean, we see ransomware gangs and cyber criminals organizing and operating like well-oiled machines at this point just the same as the defenders do. Brandon: Yeah. It's gotten crazy and you know, uh encryption comes to mind and how if you're not building powerful passwords, people are just, they're spinning their whole day trying to figure out how to break the puzzle, right? The people love to rip puzzles apart and I think it's human nature and those people know how to utilize that in the security space. And so, I love that you brought that up and so I actually have some notes I took here to actually talk about that very thing. The overall threat landscape in healthcare, I think is is your all's top target. I think it's it's ours at Philips as well. In some of the case studies that I did for this call, uh, for this podcast was, um, you know, I discovered that in 25, 26 there was a breach at ManageMyHealth patient portal exposed hundreds of thousands of medical records. It triggered legal and regulatory problems, right? We're, that's the first thing that I'm, I'm going to be worried about as a QMS guy who's, you know, um, where's my legal prospects? Where's regulatory going to come in and say, uh, we got to fix this. Because someone's accountable, everybody's accountable to somebody if you don't have, it's, like you said, the wild west, if you don't have accountability. And so, um, multiple large healthcare systems in the last couple of years have had huge problems uh with ransomware, unauthorized access and incidents, uh, security incidents last 18 months, exposing millions of records. You can see there's a lot at stake because our patient safety and quality, we call PSQ here, is a huge factor for us and how we operate, and so security is, is paramount, uh, as they say. So, you know, I don't think it's just that. The key takeaways here are that healthcare data is highly monetizable, it's operationally critical, making security failures are are not just a privacy issue, but systematic patient safety risks are included, and I think the landscape is only getting broader for that. I think, um, if we're not investing in people like Bluego, you know, if we're not investing in our, in our third party, if we're, you know, we're farming that out, if we're a large corporate company and we don't have that peace, we're making sure that, uh, that has to be part of the risk analysis we do in project management. You know, I'll stop here for a second, but I really could give you a lot of examples of where we have learned the lessons in Philips and that we're not just reactionary, we are proactive in security. Trevor: I love what you mentioned there with the patient safety. It's so important in the medical space and it's something that coming out of really most other cyber security areas and a traditional IT cyber security setup, even dealing with otherwise really sensitive systems, financial data, legal information, you know, IP protection, you don't have this added layer of patient safety that is present in a space such as health care and it puts a little bit of a twist on cyber security and it puts a twist on the way that I think we need to handle and assess risk. It's part of what can make this such a murky difficult process at times is because it's such a divergence from what we'd see as a typical cyber security and assessment. I often tell people what you might expect to normally be a critical could come down to a low and the inverse is just true. It's a little bit of a changed lens that we're really approaching it from here. Christian: Well I think in the military cyber security can cause some issues too, right? It's a little bit different than stealing credit card information. I mean, I therefore while I did defense contracting work where we have to make sure like military drones were secure from somebody hacking into them and dropping bombs on the wrong people and things like that as well. Brandon: Yeah, that's huge. You spend a lot of time in the healthcare space is kind of how we met. Is that a target market for you? Christian: Yeah, our target market is medical device manufacturers. We do some with digital health care, but primarily medical device manufacturers. We work with them to design cybersecurity into their product and to do their pre-market submission. So we do all the documentation, all the testing, uh, everything required for a pre-market submission. Then once the device is on the market, we do post-market management, which is that annual contract and continuous monitoring. So we work basically from the total product life cycle from design to disposal ideally with a client, a medical device manufacturer. Brandon: Yeah, I'm trying to kind of figure where I should leave my conversation. I think, a lot with Informatics and HL7, DiCOM. You're dealing with, if you're building a product for a customer, right? Especially an imaging equipment where Philips is, you know, in my opinion, the best at it. Of course, I'm going to say that, but um, you know, when it comes to imaging equipment, we also have to then also work with the GEs and the Siemens and the other machines that that put out a DiCom, right? That standardization of HL7 and patient records and things like that nature. Uh, as soon as you open that that gate up, right, there's a flood gate of data on patient records. There's a flood gate of, so if you're dealing with like DiCom and patient records, you know, you have to put them into a an informatic system. Do you build that into products that you're working with? Are you dealing with those kind of things? Trevor: It can be all over the place. So, you know, obviously everyone's going to have a unique requirement. DiCom is an especially tricky one to figure out, uh, the, the image format. Christian: It’s a super old protocol, isn’t it? Trevor: It's super old. I mean the format hasn't been touched since, yes, 1997 was the last time they did much with it. And since then, you've still been able to roll up malicious executable into DiCom images, and you've still been able to execute them within DiCom images. And, you know, DiCom is an, transferring DiCom images partially because it's so old, does not by default include TLS, where while passing patient records along. So it is a complicated one to figure out. When we're talking with our clients that, you know, are facing these problems, there are a couple things that we need to figure out here. The first is what can we control, and the second is what can the healthcare delivery organization, hospital, wherever it's going to be, control. And that governance question is one without really easy answer. Who is responsible for what? Should the hospital be required to enforce TLS on, you know, any DiCom communications? Generally, that's where the responsible, responsibility's going to lie right now, or should this be something where the manufacturers are making a shift and the healthcare delivery organizations have to adapt to it? And of course, this is just picking on DiCom as a specific example, but these are the types of thought paths that we have to go down when we're trying to figure out how can we solve a problem that is technically difficult to tackle here? Christian: Well, that's what's about is solving these challenging problems. Brandon: Yeah, I'd like to give you an example, I think, of what we've seen and and, uh, to get your input on it. You know, without speaking to any vendor specific or any third party vendor or anybody we've we've kind of worked with, um, here as a, you know, on my own account from my own work experience and not necessarily representing any any company at this point uh, in this conversation, but I do would like to share some of the things I think that maybe our doctors, maybe our nurses, maybe people are dealing with the actual data and they're not IT people. You know, they might have certifications in Informatics or some of the things that I've gotten smart on in the last couple of years. I know that there's a lot of certifications you get for Informatics professionalism and that full understanding of those who are listening to this right now, I think what was exciting was as I started to learn about the product because I'm coming from IT and then learning healthcare systems, you know, get smart on what an HL7 ADT is, which is your, you know, admission, discharge and transfer uh, protocols and there a lot of them aren't native be monitor or protected, they don't have encryption built in. And there's opportunity for middle software engineers like, like our companies, Philips and Bluego, that uh, we we can create those Dicon harden those HL7 and Dicon endpoints, uh, encrypt traffic. Uh, take care of our customers but still get everything working. And I can give you an example of when we go into a project and Philips is very good about doing an action report. So when we learn from it, we're not making this, you know, I like to think of Philips is like the Apple phone. It's this incredibly polished product. We go behind the scenes, we take lessons learned, we make it pretty, we bring the customer a top notch product and then we partner with people like not directly with you guys, people like you who are doing the work you're doing. So we build DMZs, firewalls. We build those things into our projects so that when we get to the project level, if you're not getting ahead of that at the SNST level, which is our service and delivery model, sales is going to sell you a product that you're going to need to have all that back end understood. So you need to have good quality people and I think from the the research I did that's not a lot of things that are taken into account right now. We need to have that built into our project planning, our program management because as we get it deeper and deeper into security issues, we're finding that those are reactionary things. We're not doing them proactively, especially security. Um, because it may cost too much. We'll get there when when we get there. But that to me is a uh two steps forward, one step back approach because now you're spending more money planning, more money waiting. I'm telling you four to six months we waited for a DMZ and a firewall to be installed because it wasn't in planning. And those those were things that were at the time not known because we thought those would be taken taken care of. But then when we get to the planning and you plug the switch in and you're the project manager there on site with your your TC team, your technical consultants are like, hey, it just red flags the the the port. Now port security is in play and the entire security network that was based behind that was not brought into the planning. and they're like, no, we're going to have to DMZ this, we're going to have to put a firewall behind it and we got to do that at 13 sites with these projects that we're doing. So now our customers are waiting and it's not necessarily a Philip's problem, but because we didn't plan for it, everybody's paying the price. And that's why we need to have knowledgeable people like you coordinated and we need to have people like you guys that are already on the books to get this ready to go. And if we're not doing that, we're wasting money and time. It's a huge part of why these regulatory authorities are pushing for the total product life cycle consideration. They they don't want to see cybersecurity treated as a point in time study, which I think is one of the bigger misconceptions within the industry is that you can say, well, when do we do cybersecurity? There is no when, it's always. It's baked into the product by, you know, design. Yeah. I think that, you know, of course we're coming from a very narrow lens here with cybersecurity. We aren't looking at the entire product and the entire development that goes into this. There's so many missing factors here that I won't speak to since I'm not going to be, you know, the the right knowledge source for that. Not the right KOL as we were talking about on our last episode. But I think that, yeah, in general, wherever you're going to need an expert in that human touch, AI's still getting there, but I think that it's it's bridging most of these gaps pretty fast. I think, uh, I'll be I'll follow Trevor in what you just said is um one of my clinicians came back to me and she's actually a peer of mine, she's a registered nurse and she works with the innovation office and one of the things we put out there was a um in our hospital patient, we have a little a little figure. It looks like almost like the muffin man, the little cookie that looks like a man with you know, just a brain man. And it says if he's having a heart problem, there's there's going to be a heart notification. If he's having a, you know, if there's something at the connection in his arm, it's going to notify the arms blinking. It's going to tell the individual, like you said, who's trying to get more time with their patients, but also get them fixed and move on to the next patient. If you have 100-bed hospital, there's only so many people you can utilize at a time to take care of all this patient, especially in ICU. So getting in there and knowing exactly where on the patient there's a problem, that AI is actually helping us in that. But what you said, Trevor, it's not going to replace the nurse. She still has to be there to fix the problem. She's still has to be there to assess whether that is the problem. And as use cases get built out, maybe it gets better in the future. We're now using, we're still using a human element to make sure that the processes and the safety and quality are where they need to be. And I think that we need to be listening to our people on the ground and not just leading so heavily into AI that we start missing the whole point of it in the first place is to build in patient safety and quality. I think the industry. A great point, right? With the Wemo is a good example of what might happen in healthcare or in a use case where you're trying to do this for another project. You asked, you know, why do things take six months? Well, now, think about this for a second. Now Wemo is out there automated. And now it's holding up a construction truck and the construction site is spending millions of dollars to get concrete poured or whatever they're doing. Now it's holding up traffic and you waste all this time on these things that you automated and then now you have secondary trigger effects to it and you're causing some of the company a problem. I mean, it's it's almost like a a perfect kind of puzzle piece that you can kind of break out and say, hey, look how messed up this can be. What do you do in that case? Like you said, you add human intervention. I think it shouldn't be that you cut all the intervention out that you build some in. And so you automate as much as you can, but you still have an oversight and you still have the ability to to make those decision makers be aware of the things that aren't for seen yet, right? You're still programming this thing. What I believe has happening right now in the workforce. Now, I'm overhead. I'm in the workforce innovation. I don't deal directly with customer. I deal with people who deal with customer. And so, you know, when it comes to how much time does a company put into their workforce innovation? How much time does companies build in risk management, risk mitigation. That's kind of where I'm at. I'm in there to reset the ship when we need to and I'm building out processes. So how do I better and then get rid of some of the overhead where I know that I can save the company money, get a better product for the customer, but I still am bringing myself into the value stream. If I do so well that I cut myself out of the value stream, there's, that's not really intuitive for a person who's career-minded. What I want to do is I want to get so good at AI that I'mble. I'm a problem solver. I'm someone who's utilizing it, not scared of AI, but I'm someone who's going to go grasp it. I'm going to start reading it as much as I can. I'm going to get on Chat GPT and get an account. The company may not even pay for that, but I'd taken the first time that I got into Philips, I started doing AI power core presentations. so I wouldn't waste time on. I started taking Chat GPT into account where I could that didn't have anytion information in it. and what could I streamline? A lot of our stuff is in how do we sell the product? How do we just, you know, get it on prim? How do we then secure it? How do we build a project around it? And now it's becoming program management for me because now I'm also trying to service it with service Max and some of the other things that we have at play. What can I do right now to help all those people that are right now scrambling to get the right information in the right timing so that all that orchestrates for our biggest customers. We have a lot of big hospitals obviously at Philips. But one of the biggest things for me and we we hit it on at the beginning of the conversation is the VA. The VA, I'm a VA member. Whatever benefits us to get things right is going to substantially make it better for myself and my peers. because we're all part of that VA system coming out of retirement, coming out of serving four to six years, whoever served, you know, thank you for your service. We want to get those things right at the VA too. How do we data stream? How do we collect, build work streams that are deliberate and fixing problems. And one of the things we're doing right now is using AI and this initiative I've taken is I want to use AI to get everybody's ideas, everybody's runbooks, everybody's processes, whether it be clinical, technical, CPM, whether it be sales, whether it be all of these different people who have all this work on their plate and they're dealing with customer I know I call them and I can get barely in 15 minutes to get a conversation with them just to find out what's not working. and they're so busy, I don't even get an edge a word in edgewise. That breaks my heart because I'm like they don't even have time to give us time for process of improvement. And what I try to do is utilize AI to grab all of the workbooks that they have on. Everybody has their own thing. The company, why does it take so long to get stuff done is because the human factors of things change very dynamically. And so what one team is doing isn't always in stride with the next. What I'm utilizing right now is building in AI to grab all of these workbooks, align them and a work breakout structure and a process that will help our project managers and our sales team know exactly on the clock where we're at with the project. These are initiatives that are first of kind and we have to test them. We have to have time to make sure everything's doing what's supposed to do. Otherwise we're going to be sitting on a bus not know what to do. We can't get around the bus because we're way out. So, you know, I'm trying to build out the the idea that we can get around the waymo. We can uh fix the way I don't know if we should use as the new the new word for like not moving forward is uh we're we're trying not to be the waymo right now. and it's no crash against them. They're working like we are to try to build build fix actions. but um that's what I I see. Look, if you're listening to this and you're wondering how you can develop your career, what you're want to get into Mettech, be a problem solver and grab a hole of AI because it ain't going nowhere. It's only going to get bigger and you can future proof yourself and your teams by being a problem solver. Don't be afraid of AI, you know, be behind it. And I think that's the thing we need, I think help people stop the fear hungry. AI is not there to take our jobs. It's there to make us more efficient and to drive uh what our customers need. That's they need our time. So, if AI can help us do that, I think our conversation can go down that road. Awesome. Uh, well, it's what we're coming up on time here. I'd like to go around the room and see uh everyone has the parting words or last minute words of wisdom. So I usually start with Trevor and uh I expect you to say something different this time Trevor. Hey, I've been pretty good about different the past. I'll give it like the past eight or nine. So, I think something really important is we were touching up on edge cases. Edge cases are always going to be something especially in the era of AI that are hard to predict and they're hard to control. In the medical space, these edge cases and like you're saying, Brandon, having this human element as a check point inserted into the system is increasingly valuable. The more we're handing over to AI, the more it's going to be missing some of these edge cases. And so we were just the other day talking with a cardiac surgeon who's saying, you know, hey, I'm getting all this input, I'm getting all this information from all of these different machines and noises and beams and alarms and signals. The patient is what I'm using as my main, you know, is something wrong. If these alarms are saying, you know, oh, there's a crisis, the patient's bleeding, but I look at the patient, the patient's not bleeding while I'm operating on them, blood pressure stable. Maybe these alarms aren't giving me the right. They're just creating noise. They're not giving me signal here. So, understand your edge cases and understand how to handle them in an increasingly automated world. I think, you know, of course, we're coming in from a very narrow lens here with cyber security. We aren't looking at the entire product and the entire development that goes into this. There's so many missing factors here that I won't speak too since I'm not going to be, you know, the the right knowledge source for that. Not the right KOL as we were talking about on our last episode. But I think that, yeah, in general, wherever you're going to need an expert in that human touch, AI's still getting there, but I think that it's it's bridging most of these gaps pretty fast. How do you train a I for all these edge cases though? I think this is where the problems come up. Like you Trevor, you talk about Wamo, we talk about Wamo a lot. And there's Wamo here, but you're saying a human has to take over the Wamo if it's like certain traffic conditions. So it's like a Wamo has been out a long time. Like why haven't they figured this out? There are too many unique. Yeah, there's all sorts of weird stuff. You know, they're paying people to walk around and close the doors for Waymos now because they haven't figured out how to automate the door closing process. A couple days ago, I was taking a waymo and there was one of these huge uh like a greyhound bus just parked in the middle of the street. And this is in downtown San Francisco, narrow street, you know, there's kind of just chaos all the time going on. And so we get behind it and the cars are going around the Waymo, around the greyhound bus and the Waymo keeps going into reverse, backing up a little bit, going forward, backing up a little bit, going forward. It's freaking out, it doesn't know what to do. Thinking back on it, I've never been behind a greyhound bus in downtown San Francisco. So if I were a machine, I was not trained on what to do in that situation. I can connect the dots and go, I should probably make sure there's no car in the other lane and drive around the greyhound bus. But the waymo wasn't connecting the dots, and so it had the little thing that pops up, you know, hey, we're getting it back on track, someone's assisting it to get out of here or whatever. But it when we have these unique situations, that's why AI is only going to be as good as what you've trained on. It's garbage in, garbage out, and if you're putting quality training in, you're going to get quality output out of it. If you don't have any training on a situation, you're probably not going to get any output from it and your isn't going to just go back and forth in the middle of the street. Well, with a with medical devices too, the the risk is very high if the device isn't trained from AI perspective on specific use cases, right? It's not like I mean it's high in traffic too, but the way is going to air on the side of not doing much like that situation. Like I had a similar situation here in Phoenix there was behind a construction truck in it, it had a blinker going both ways. So the Waymo kept turning this way and this way and this way like for like 10 minutes. I think finally somebody took it over. The industry whatever you bring up you're with the Waymo is a good example of what might happen in healthcare or in a use case where you're trying to do this for another project. You asked, you know, why do things take six months? Well, now, think about this for a second. Now Waymo is out there automated. and now it's holding up a construction truck and the construction side is spending millions of dollars to get concrete poured or whatever they're doing. Now it's holding up traffic and you waste all this time on these things that you automated and then now you have secondary tertiary effects to it and you're causing some other company a problem. I mean, it's it's almost like a a perfect kind of puzzle piece that you can kind of break out and say, hey, look how messed up this can be. What do you do in that case? Like you said, you add human intervention. I think it shouldn't be that you cut all the intervention out that you build some in. And so you automate as much as you can, but you still have an oversight and you still have the ability to to make those decision makers be aware of the things that aren't for seen yet, right? You're still programming this thing. What I believe has happening right now in the workforce. Now, I'm overhead. I'm in the workforce innovation. I don't deal directly with customer. I deal with people who deal with customer. And so, you know, when it comes to how much time does a company put into their workforce innovation? How much time does companies build in risk management, risk mitigation. That's kind of where I'm at. I'm in there to reset the ship when we need to and I'm building out processes. So how do I get better and then get rid of some of the overhead where I know that I can save the company money, get a better product for the customer, but I still am bringing myself into the value stream. If I do so well that I cut myself out of the value stream, there's, that's not really intuitive for a person who's career-minded. What I want to do is I want to get so good at AI that I'mble. I'm a problem solver. I'm someone who's utilizing it, not scared of AI, but I'm someone who's going to go grab it. I'm going to start reading it as much as I can. I'm going to get on Chat GPT and get an account. The company may not even pay for that, but I'd taken the first time that I got into Philips, I started doing AI powerport presentations. so I wouldn't waste time on. I started taking Chat GPT into account where I could that didn't have any feature information in it. And what could I streamline? A lot of our stuff is in how do we sell the product? How do we just, you know, get it on prim? How do we then secure it? How do we build a project around it? And now it's becoming program management for me because now I'm also trying to service it with service Max and some of the other things that we have at play. What can I do right now to help all those people that are right now scrambling to get the right information in the right timing so that all that orchestrates for our biggest customers. We have a lot of big hospitals obviously at Philips. But one of the biggest things for me and we we hit it on at the beginning of the conversation is the VA. The VA, I'm a VA member. Whatever benefits us to get things right is going to substantially make it better for myself and my peers. because we're all part of that VA system coming out of retirement, coming out of serving four to six years, whoever served, you know, thank you for your service. We want to get those things right at the VA too. How do we data stream? How do we collect, build work streams that are deliberate and fixing problems. And one of the things we're doing right now is using AI and this initiative I've taken is I want to use AI to get everybody's ideas, everybody's runbooks, everybody's processes, whether it be clinical, technical, CPM, whether it be sales, whether it be all of these different people who have all this work on their plate and they're dealing with customer I know I call them and I can get barely in 15 minutes to get a conversation with them just to find out what's not working. and they're so busy, I don't even get an edge a word in edgewise. That breaks my heart because I'm like they don't even have time to give us time for process of improvement. And what I try to do is utilize AI to grab all of the workbooks that they have on. Everybody has their own thing. The company, why does it take so long to get stuff done is because the human factors of things change very dynamically. And so what one team is doing isn't always in stride with the next. What I'm utilizing right now is building in AI to grab all of these workbooks, align them and a work breakout structure and a process that will help our project managers and our sales team know exactly on the clock where we're at with the project. These are initiatives that are first of all and we have to test them. We have to have time to make sure everything's doing what's supposed to do. Otherwise we're going to be sitting on a bus not know what to do. We can't get around the bus because we're way out. So, you know, I'm trying to build out the the idea that we can get around the waymo. We can uh fix the way I don't know if she uses as the new the new word for like not moving forward is uh we're we're trying not to be the waymo right now. and it's no crash against them. They're working like we are to try to build build fix actions. but um that's what I I see. Look, if you're listening to this and you're wondering how you can develop your career, what you're want to get into Mettech, be a problem solver and grab a hole of AI because it ain't going nowhere. It's only going to get bigger and you can future proof yourself and your teams by being a problem solver. Don't be afraid of AI, you know, be behind it. And I think that's the thing we need, I think help people stop the fear hungry. AI is not there to take our jobs. It's there to make us more efficient and to drive uh what our customers need. That's they need our time. So, if AI can help us do that, I'm sure our conversation can go down that road. Awesome. Uh, well, it's what we're coming up on time here. I'd like to go around the room and see uh everyone has the parting words or last minute words of wisdom. So I usually start with Trevor and uh I expect you to say something different this time Trevor. Hey, I've been pretty good about different the past. I'll give it like the past eight or nine. So, I think something really important is we were touching up on edge cases. Edge cases are always going to be something especially in the era of AI that are hard to predict and they're hard to control. In the medical space, these edge cases and like you're saying, Brandon, having this human element as a check point inserted into the system is increasingly valuable. The more we're handing over to AI, the more it's going to be missing some of these edge cases. And so we were just the other day talking with a cardiac surgeon who's saying, you know, hey, I'm getting all this input, I'm getting all this information from all of these different machines and noises and beams and alarms and signals. The patient is what I'm using as my main, you know, is something wrong. If these alarms are saying, you know, oh, there's a crisis, the patient's bleeding, but I look at the patient, the patient's not bleeding while I'm operating on them, blood pressure stable. Maybe these alarms aren't giving me the right. They're just creating noise. They're not giving me signal here. So, understand your edge cases and understand how to handle them in an increasingly automated world. I think, uh, I'll be I'll follow Trevor in what you just said is um one of my clinicians came back to me and she's actually a peer of mine, she's a registered nurse and she works with the information office and one of the things we put out there was a um in our hospital patient monitoring, we have a little a little figure. It looks like almost like the muffin man, the little cookie that looks like a man with you know, just a brain man. And it says if he's having a heart problem. There's there's going to be a heart notification. If he's having a, you know, if there's something at the connection in his arm, it's going to notify the arms blinking. It's going to tell the individual, like you said, who's trying to get more time with their patients, but also get them fix and move on to the next patient. If you have a 100-bed hospital, there's only so many people you can utilize at a time to take care of all this patient, especially in ICU. So getting in there and knowing exactly where on the patient there's a problem, that AI is actually helping us in that. But what you said, Trevor, it's not going to replace the nurse. She still has to be there to fix the problem. She's still has to be there to assess whether that is the problem. And as UCs is getting built out, maybe it gets better in the future. We're now using, we're still using a human element to make sure that the processes and the safety and quality are where they need to be. And I think that we need to be listening to our people on the ground and not just leading so heavily into AI that we start missing the whole point of it in the first place is to to build in patient safety and quality. I will add something that you said Brandon that I think is extremely important for anybody looking for a career or a pivot or a start a company is to find a way to add value like you mentioned. And then don't be afraid of AI. I hear a lot of times people are afraid of AI taking their job. But if you can learn to use AI to your benefit, learn to prompt it properly, set up your custom GPT, whatever you need to do, it will only help you and it will help you add more value to whoever you're trying to add the value to as well.