Episode 2: One Row Is a Patient, with Dilip Raghunathan

Final transcript, lightly edited for readability. Timestamps match the published episode. The views expressed by guests are their own.

00:10 Patrick Coyle: Hi everyone. I’m Patrick Coyle. I created Patients + Profitability because I spent my career seeing what happens when patient access and healthcare economics become disconnected. Most organizations manage those separately. I believe they’re actually the same conversation. When patients struggle to access therapy, businesses struggle too. My work focuses on helping organizations better connect patient outcomes with financial sustainability. Every episode, I sit down with one person who’s lived both halves of that sentence. I want to introduce you today to Dilip Raghunathan. Dilip will tell you more about himself.

00:50 Dilip Raghunathan: Hey guys, name is Dilip Raghunathan. I’m a immigrant to the United States since 2001. Then went to high school here. Biomedical engineering graduate from Rutgers University, really got deep into biology and biological sciences at that point. Got out of that and went into the industry as a statistical programmer. Got my hands with software programming and data engineering and analytics and tools and so on and so forth. And very quickly found myself into things like governance and data standards and stuff, and and once I did that then for more than a decade as an entrepreneur, I quickly realized that the pressure points and the pain points of companies all comes boiling down to just one thing, right? Which is understanding the ability for a company to get into change management and then successfully navigating it.

00:50 And then the topic could be anything. It could be that they want to adopt governance, they want to adopt standards, they want to adopt new technology. So I spent a bunch of time doing that, and I then went on leading teams in stat programming, data governance, as well as data standards, CDISC standards, and finally I found my home at Regeneron, where currently I am now the solutions partner, and I lead the portfolios of very data-heavy business units such as biostats, data management, stat programming, clinical sciences, clinical imaging, precision medicines, translational sciences. So I have a lot of fun doing everything that I did as part of the business. Bring the rigor of consultants and consultancies on how to approach a problem and how to take that problem all the way from a concept to a solution.

00:50 And more than anything, I’ve been blessed to be at a part where I have both the business as well as technical acumen, and to be born in the age of AI. And that has extremely excited me over the last few years, and I see nothing but just opportunities going out into the future. So glad to be here, Patrick, and thanks for finding me out of the ether and bringing me here for your podcast.

02:53 Patrick Coyle: Yeah, thanks for joining us, Dilip. It’s so important for us to get this back out into our audience because there’s so much going on. When you when you and I first talked, we covered enormous amount of ground, but we kept circling back on one question, and that’s what I want to spend our time on. You’ve had 25 years on one side of the approval line. I’ve spent nearly as much on the other side of the approval line. But we’re both watching the same thing fail in same four ways: you and R and D and IT, me and commercial and patient access. Nobody’s connected those two observations, as far as I can tell. So that’s what we can do today. So at the beginning, what pulled you into this type of work, and what made you stay?

03:34 Dilip Raghunathan: Okay, there’s there’s an answer to that, which maybe most people would just chuckle at. It’s frankly just fate, right? And I keep saying this to a lot of people. I walk into a room. I’m in a meeting with, say, for example, recently I was also talking to somebody about patient access, and as you just rightfully said, I spent 25 years of my life on the other side, which is the discovery and the development of drugs and so on and so forth. And I was talking to this person about the nuances of affordability and patient access and the way that you do secondary insurance and the way that you do afford patient affordability programs and market access and how the primary and the secondary insurances work and how the copay works and all that stuff.

03:34 And they were staring at me, going, “How the hell do you know all this? And I said, my wife works in that field, so I picked up a ton over a decade that she’s been doing that. Right. So the point is that I think that what got me into this field was essentially almost seemingly a random series of events, by going from biomedical engineer to getting into computer science, computer programming was in and of itself a bit of a shift that I did not anticipate. It just so happened, and it was easier to take that because it was the first job coming out of college. And hey, there’s a job, there’s money. Why not? Let’s go, right? And biomedical engineering to pharmaceutical companies. It did not seem too much of a.

03:34 Of a shift later, when I did my master’s, I did make a conscious decision of making my master’s a dual sort of a program, which is information sciences as well as business, and that really solidified the fact that I’m going to be married more to technology and business around the technology rather than the biology of things, and that sort of then catapulted me to the later part of my careers, where I still kept myself right within the clinical development phase, but instead of being in the business and handling portfolios better, required technology. So it was always just edging towards technology to handle the portfolios better, and then I just became the tech lead and the tech liaison, and now I’m essentially an IT leader representing the business interests, taking business strategy, and then making an IT strategy in order to jive with that, and then create success after that. In a way, yeah, random, but I believe that I’ve just been able to put together the pieces of the puzzle in a way where I can use all of that for the current role that I’m in.

06:04 Patrick Coyle: It’s amazing. One of the things that jumped out to me what you just said is your life partner helped in that evolution. Right, so often our home life is completely separated from our work life, but in reality, we’re constantly evolving, and that starts at home, and then it continues at work, and then before long, it begets its own cycle. But before we go any further, I wanted to say everything you say today is your own capacity. It’s still personal capacity. I’m not asking how Regeneron does anything, Regeneron is where you work. I’m asking about the industry. Totally different questions. So just wanted to put that out there. When did a row in a data set stop being just a row for you?

06:46 Dilip Raghunathan: That’s a great question, right? Let me put it this way: coming out of college, I already had decent amount of computer science courses and all that stuff, right? So I knew data. I knew that you take some data, whether you enter it in a tab-separated format or a comma-delimited format or in an Excel spreadsheet, and then you import that in. I knew that data, the way data is always columnar, right? It’s tabular, it’s two by two, like it’s rows and columns. Done, right? Then obviously columns and what they are means something, and each row is a thing in that data set, and it means something. When you get into clinical trial programming, and I used to use this language called SAS in order to do a lot of my coding back in the day, they teach you about like programming data vectors and how like the the underlying computer science language really accesses the data, and then how that gets compiled, and how it really interacts with that data.

06:46 And in that, there are things that act upon columns as a whole. There are things that act upon row by rows, and there are also things that act upon cells. Each one of those cells within that data, in a weird way, learning how to program and code to call out a cell or a column or a row or the entire dataset as a whole is when I reinforce the distinction between that. Now, that’s physical distinction, right? What does that rule actually mean? If you’re looking at a phase three oncology trial and you’re looking at a data set of all the subjects who are enrolled in that trial, you’re just pulling up a data set, you’re crunching numbers, and you’re saying my count is I count 1251 rows, and you tell the other person, and the other person’s going, It’s 1252 rows, and you’re like, ah, close enough.

06:46 That one row is a patient that could be your family member, that could have been your own grandma, your own wife, your own husband, your own father, son, anybody. Right? The meaning of a single row of data could be the difference, all the difference in the world to a bunch of people in the world, right? So, when you work on clinical trial data sets, and I know that you’re you hold that dear to your heart: patients, the privacy, the security, the governance around things and so on and so forth, it elevates what a data point means because that data point or that row or that column or that data set represents a lot more than just a piece of data, and you wouldn’t really maybe find this sort of in maybe other industries if you’re developing a cell phone app to do Candy Crush, right?

06:46 A data set means maybe a lot of people’s interactions, clicks, and what they like, and where they go on the screen, or where their eyeballs are, and all that stuff. Great, it’s all important stuff. At the end of the day, we got to play Candy Crush and have fun too after doing all the work. But in life sciences, a row of data really means a lot more than just a row of data. It means humans. That’s why it’s really very important.

10:03 Patrick Coyle: Wow, that’s all I could say. In our business, data represents the visualization of a life lit on a treatment journey, on a prescription journey. Absolutely fascinating. So, if you think of that responsibility that we have, and for me, that that aha moment was when I started looking at copay economics and fall off for a patient. It wasn’t. I didn’t have a journey in mind. It was a data point. Look at this disparity in the data. This seems like a trend. I didn’t know about commercial copay at the time. I was new into U.S. business, but I found that there’s a pattern here, and this is way before all the things everyone talks about today. So very similar for you. I turn that into how can we leverage that information and improve an outcome, right?

10:03 And that journey is never the same in the real world as it is in the clinical world. But if we take that and we zoom out now and scale it, you’re not responsible for one person’s data anymore, right? You’re responsible for systems that hold hundreds of 1000s and millions of them. So, what changes about the job when it goes from one to a billion, and what has to stay the same throughout?

11:14 Dilip Raghunathan: Great question. So, what changes the magnitude, the scale? You’re dealing with a lot of patient data, and you’re dealing with the responsibility and perhaps even the accountability of now not looking at the individual, but also looking at the whole. Therefore, the trends. Therefore, the signals. Whether that signal and the trend is pointing towards a drug being efficacious, or whether that signal being pointing to a drug safety, maybe it’s a safety concern. Maybe it is a it’s a great outcome for a safety because you’re seeing a signal in a patient population where certain adverse events are not happening as much as it used to happen in that patient population before. Whatever that is, you’re responsible for the whole now, and with the whole comes the responsibility to analyze it as a whole, and take make intuitive judgments about what those signals could be, but then define the scientific and statistical rigor, having had some statistical training in my past, is very important to make sure that you don’t misinterpret a signal or assign causation for somewhere where there’s only correlation, right, and so on and so forth.

11:14 So you do have to be responsible and accountable to the whole, you have to be responsible to search and be curious and inquire. And when you find something, you cannot be very judgmental. You have to take it objectively, and then you have to use the rigorous statistical sciences in order to prove or disprove your hypotheses. Right? That’s all that is different, or it’s more when you have the whole as opposed to the the one, but what doesn’t change is that you’re still dealing with patient data. You’re still dealing with 100 billion records of patient data, as you said rightfully. It represents a journey of a copay, a journey of prescription medications, and so on and so forth. It is things that are personal. It is things that depict humans, and it’s their contribution.

11:14 What I analyze, and I’ve done this before, I’ve taken a lab dataset, like a lab values results dataset, from across the programs within a somewhat of a large pharma. We had a file that was about close to about one terabyte huge, and it had maybe 21 million records or so at that point, right? When you just stop and think about what those 21 million records mean, it meant a whole lot of blood, just blood, right? From these patients, there was enough blood drawn across all these patients that now you were able to take that blood, spin it. A bunch of people were put put to work for that, right? As their contributions as well, and then you get all these results of analytes in within your blood.

11:14 And now over a period of time, for all these people who individually represent a different walk of life, and whether that is as simple as they were recruited from here versus there, Europe versus U.S. or 19 years old versus 85 years old, or gender, and so on and so forth, or the race of where the ancestors came from, and so on and so forth. Whatever it represents, it gives you all of that information along with their blood, which then gives you the information inside of their body. Imagine the amount of personal sort of info that you are holding. At the same time, the amount of contributions that these folks have done for you to get that the personal info. Right. So it does not change your responsibility to keep that data.

11:14 Governed to keep that data secure, and once again to make right use of that data. Right in in the business of where data is the oil, resellers are ubiquitous. Right, take data, clean something, just put it over the other side, resell, resell until it get to a final scientist where the scientist is actually using that data for a particular intended purpose. Yes, there’s always middlemen to bring the data about. It’s fine, but every middleman, everybody who holds that data, has to understand where that data came from. It came from people, and in terms of biological data, it came from literally blood, sweat, and tears. Right, so we got to hold it there.

15:37 Patrick Coyle: Direct correlation, something I almost never talk about, is I spent almost a decade in IT solutions, and I started in financial operations, and then it pivoted into solutions that Computer Horizons was creating for pharma clients and just industry clients around Y2K. So I’ll totally date myself. But what I learned was all about the preservation of data and data structure, well before we’re anywhere near talking about these things. But so when I got into pharma, I immediately saw that potential that you were living all the time. Let me ask you about Alfred Mann, a man who already made his money in avionics, picking up inhaled insulin because he wanted to save the world. This is legacy, and you helping bring it to market. Tell me about that.

16:25 Dilip Raghunathan: It was one of the. I was frankly I was just blessed to be part of that journey. It was this was back in 2008-ish where I was brought into this team, which was the company was named MannKind Corporation after Alfred Mann, Mann with two Ns, and he was a legend already. And going to the company, I did not know about him before going to the company. But I went into the company. It was a. It was supposed to be like a very short contract. I was brought in to do a, hey, come here. We’re doing a submission of a drug, and your responsibility is going to be create data standards, make sure that all all of the data sets are in CDISC particular format, which is acceptable by the agency.

16:25 We want to fast track this because we really believe that our review times from the agency with the PDUFA will reduce if we do it in the standard. Great. So I just walked in there almost unknowingly, right? What this whole thing is about? Yeah, I knew that they were in diabetes, right? That’s pretty much what I did. I went there, and then I learned about Alfred Mann. I learned about his life, his legacy, the whole nine yards, and then I learned about what we’re actually doing. Cool signs. We took a powder. It was called a technosphere inhalation powder. We took the powder. We put human insulin in inside that powder, like basically attached it somehow. That was a cool site. The scientists figured out how does it attach and how does it stay stable, right?

16:25 And once that is done, instead of having that whole thing bottled up into a pill and swallowing it, the whole idea was if you are able to figure out a way where insulin disperses into the body very quickly, which is how the subcutaneous insulin goes in, right through your bloodstream. If you’re able to get it in a way where it’s non-invasive, it was going to be game changing. When you say non-invasive, and when you say medication, people always go, “Oh, medication! Like somewhere through the mouth, right? Okay, but the scientists here figured out that it was going to be a inhaled insulin. When we did that inhaled insulin, it mimicked the natural response of the body after having food, and your body recognizes that you just put some glucose in there, so your body Releases insulin, right?

16:25 The response time that your body takes to have the food, know how much food you’ve had, know how much sugars it’s got in, and then put the insulin in your body, was extremely similar to this inhaled insulin, and it was a very cool device. It was like a whistle. You put a little cartridge in there. It’s self-titrated cartridges like five IU, 10 IU, 15 IU, and you can do combinations thereof. Why? User experience. Because you could go to the restaurant and you could order some food. You start eating. You think that food’s going to be great, but one bite later, you’re like, “Oh crap! I just this is this horrible, and maybe you’re just eating a little bit here, a little bit there, and then you’re just not eating as much.

16:25 In the past, before a person sits there and eats, they go excuse us into the bathroom. They go shoot themselves with insulin, with a certain amount of IU, which is based off of how much they’re going to be eating, and therefore their glucose spikes, and they do that, and they come back. But if they make a decision not to eat as much, now they’re going into a adverse event because of the fact that they have more insulin, hyper hypoglycemia, so that kind of stuff, right? So the point is that this inhaled insulin not only changed the mode of. Delivery, which is instead of shooting, you inhaled it, but it worked on the ultimate user experience, which is just the way that you would make your normal decision.

16:25 The body would make that decision for you in terms of you eat half your plate. The certain amount of insulin is going to get developed, right? If you eat the whole plate, a certain amount, you would be able to titer your doses based on how much you actually ate, rather than how much you are maybe about to eat.

20:26 It was such a game changer. Great drug, drugs approved. It’s out in the market. We went to the FDA. We were we appeared at the Great Hall, and we had an advisory committee, and I think we won 10 two two noes and 10 yeses, and it was clear as day that this was going to be a life changer for a lot of people with diabetes. So it was just blessing to be part of that.

20:48 Patrick Coyle: I think that’s amazing, and to get to that point, and it’s very interesting. You find these little things, these nuggets as you go, and I’m pretty sure MannKind was one of the clients of Computer Horizons when I was there. Didn’t even remember that until we’re just speaking here, and the other thing you were talking about with inhaled insulin, I was thinking of inhaled amikacin, a non-tuberculous mycobacterium avium complex, right? And finding two sides of the same coin, maybe again, right? Where we’re working in different roles for the same company. Maybe tell me a little bit about your experience there.

21:20 Dilip Raghunathan: Yeah, it’s amazing because you just reminded me that I have worked on a lot of inhaled drugs. I don’t know, maybe just a coincidence from college. NTM, once again, rare disease, ultra rare, like 5000 people or so a year in the U.S. or something like that, a little bit more globally. Insmed, the company that I was working for before, they got into that space where they figured out that these patients are usually having a lot of comorbidities, so their life is miserable, anyways, and then they have this very ultra rare bacteria bug sitting in your system, which is really kind of it’s it’s exacerbating the cough. It’s it’s creating a lot of mucus within your lungs and so on and so forth because your lungs keep trying to fight it at the end of the day.

21:20 And all the antibacterial or the antibiotics that were in the market that were treating this ultra rare disease. Let me put it this way: they were there not to treat this disease. Doctors were forced to use things that are out there to treat this disease because who’s going to make a drug for 5000 people a year, right? Who’s going to profit out of that? Which is also a different problem when you talk about biotech and incentivization and so on and so forth. And a lot of patient access and the kind of programs that FDA runs and so on and so forth is trying to break that barrier down, where we’re not just focusing on your heavy hitters or your drugs, or that that can make more revenue, but we’re also trying to focus on human lives, which is very necessary, right?

21:20 So this drug was once again-it’s a very common sort of. If you ask a five-year-old, Papa has something in his lungs. It’s a bug sitting there. This bug is a very mean bug. It’s very hard to kill that bug, and the only thing that can kill that bug is an antibacterial whatever medication or antibiotic, right? But we got to eat that, or we got to put that in our blood. And by the time it gets to the lung, it becomes so weak that bug is able to take its shield and say, “Not now, drug. What will the five-year-old say, “It’ll say, Papa, why don’t you just put it directly into your lungs? That’s it. It’s as simple as that. the The bug sat inside the lungs.

21:20 The medication went inside the blood or the gut, and all we had to do there is to make sure that we’ll figure it out. First principles thinking. First principles design. Let’s figure out how to properly, safely, and in an efficacious way administer an antibiotic into the lungs. And it was a device which was a liposomal sheath and a coating, which is once again like cool stuff in the lab. They took the actual amikacin, which is nothing but an existing antibiotic, they wrapped it on a fat shield called a liposomal shield, and that fat shield wouldn’t disintegrate unless it goes into your alveoli and it disperses and so on and so forth. So it just goes straight to your lungs where your bacteria is sitting, and it’s just going to kill them out there, and then you’re going to be fine after that, right?

21:20 And we proved that with like the bacterial load decreasing and so on and so forth. FDA was a lot more concerned about forget about all the cool science stuff and how you’re killing the germ. Is the patient feeling better at the end of the day? And so we went out on seeing clinical outcomes, right, and trying to do eCOAs and ePROs and whatnot to make sure that the patient was feeling good, that they have tangible betterment of their symptoms, but once again, there was a comorbid like people are comorbid having comorbidity with other conditions. So it was like, oh, are they not feeling good because of this now, or are they not feeling good because of the COPD or asthma, right? And so on and so forth.

21:20 So we had to go through a few hurdles. There’s nothing that in life that comes without going through. But the idea was that genius science ingenuity met a relentless operational excellence. Will the CEO there? He was just gung ho. Like he knew what to do.

25:12 He knew how to do it, and we just got it done there. So it’s a great team, great experience. Once again, we appeared in the Great Hall, and we did that. Here’s that drug. Here’s why we need to be approved. We were approved, I think, unanimously, if I’m not mistaken, for that one, and and yeah, and the drug is in the well, drugs in the market now is still being prescribed the same way as MannKind’s inhaled insulin is, and and the the world is all the more better. And Insmed developed a muscle to get into rare diseases, and they replicated that, and then they also got a muscle to be successful at development of drugs, so they’ve already had some other approvals and whatnot. So the world’s a better place.

25:48 Patrick Coyle: So another brilliant insight that came out right in our first meeting was I when I left Novartis, I followed the U.S. CFO Paolo Tombesi over to be the CFO at Insmed. So while you were there working on that, I came over right after interim phase three data was positive, and then helped build for readiness. Right? Why did I do that? Why would you leave this amazing big thing opportunity that I did at Novartis? Was because I had never launched helped in a part of the commercial and financial launch of a company like that before, and that opportunity was very compelling, right? So I did that, and I got to live through that whole journey for our case. So super, super exciting. But don’t let’s not forget about for the future that the thing that you uncovered there was first of all we were on separate sides of the aisle for a couple of these times, and at the same time you said inhaled substances as a new delivery mechanism for even something old, right?

25:48 Let’s not forget that because there’s amazing stuff going on there today. I think we need more of that innovation out there in the market, but I want to turn to a totally different thing now for the rest of our time. You said something on a panel that I saw, and I’ve not been able to stop thinking about that. 95% of AI pilots in life sciences never make it to production, and that’s a living and breathing number. Who knows? Maybe, maybe it’s even more. Right? Should be less right now, but it’s not a technology problem; it’s a culture problem. I could not agree more. What? Tell me a little bit about that, and what you actually see when that pilot dies. And the correlation for me is on the advisory side. Here’s a great solution. Here’s the next gen thing. But guess what? What happens? Reluctance, right, to adopt, embrace change. Tell me about it. Yeah.

27:33 Dilip Raghunathan: So what you referred to there, I think I was in the stage in the SAS Innovate conference in Dallas, and and I was asked about why do all pilots fail, or what can leaders do better to make to get off of pilots and then go to production? I think that was a question: get off of pilots and go to production, as if pilots were so successful, and what we’re trying to do is like just get off of it and go to production. And this one, I have to remind the crowd in a very bleak way that 95% of all pilots in AI has failed, and I repeat that like two times in that crowd until it like sank in fully. Right, you’re right. It’s it’s less about technology, in my opinion.

27:33 I think with some of the years that I’ve spent in technology, just deep, and then prior to that in business and technology out there to enable us and us adopting that, and so on and so forth. It almost never was about technology, frankly, right? In terms of like these failures of pilots and so on and so forth. Like, what is a pilot? A pilot is when you say to yourself, “I’m doing something now. I think that I could be doing better, or I don’t know if I can be doing better. I just want to see if I can be doing better. Let me try something different, right? And then you try something different, and maybe the different works really great. Maybe the different doesn’t so work.

27:33 95% of AI pilots failed. I think that number is more for AI than non-AI pilots, frankly, right? It’s got to be like way more, right? Which is why it made the headlines and so on and so forth. So here’s the reasons why I think that is the case, and this is a typical thing that happens with every hype that happens in the market, whether that’s decentralized ledgers and cryptocurrencies and so on and so forth. That technology coming into to our space, whether it be cloud computing, whether it be distributed processing, and if anybody in this podcast is following tech, whatever happened to Hadoop and the Hive and things like that, right? Like it was supposed to be like the next thing in the world, and so on and so forth.

27:33 Then they kind of non-existent, or they got rolled into other things as well. But the point is that AI, the hype of AI, wasn’t just one of those nominal hypes. This was one of the biggest hypes that we’ve had in the last. I think the last hype this much was internet, but at that point, Facebook or LinkedIn or Instagram and things like that did not exist, and it wasn’t really the digital age, obviously, because internet was just being born, right? So there wasn’t a. Uncontrollable sort of like a cycle that went in, making that hype more of a hype, and that hype more of a hype, and so on and so forth. Right, the vicious cycle. With AI, we have that. Right, and first, I think it was a lot of expectations on something that was fairly new.

27:33 So, from a technology standpoint, by the time people started adopting GPTs, we were already in GPT threes and fours. GPT threes and fours can do a lot of good stuff, right? You still have to be careful about prompting it and so on and so forth. But the the point is that people expected AI to just walk in and do it all for them, and then when they realize that’s not the case, there was a little bit of the letdown or the disillusionment, which caused to some of those pilots just being deemed as a failure because the requirements or the expectations from these things were a lot more than it should have been. Number one, number two, even for places where they carefully thought through the capabilities of AI, and then matched it with a use case, and then did this, and then went ahead and proved that out, and then they said victory.

27:33 We have achieved 55% of acceleration in this workflow, and then they would take that workflow, and then they would say, okay, who’s responsible for this workflow, and how can we integrate it into your system? Pilots failed into production, right?

31:25 The pilot was maybe successful, but it never scaled into production. Why? That group, that organization, was probably never part of this thought process of can AI be used here? Should AI be used here? How do we use it here? Is it going to change the way we work? Is it going to is it going to take into consideration the more important values and the principles and so on and so forth that I have for doing my job, or is it just going to change my job in a way where I don’t agree? If it makes it easier, why not? But if it changes the job the way that I don’t, and it’s a lot of people just paint that whole thing under oh these people are disgruntled oh they just scared about their jobs and so on so forth I don’t think so right because I’ve talked to many people in the business now that I’m in this role everybody is genuinely wanting to take technology and take a leap forward right that’s how mankind has gotten from fire to agricultural wheels to industrial revolution to electricity to telecommunications to internet cloud computing and then AI right if we were all just a reluctant bunch of people we would have never gotten out of the caves in the first place right so it is very important to understand that people are not just blocking these things we need to bring about a awareness of the fact that there’s a technology, here’s its capabilities, there’s a group of people who are going to be using it for a specific purpose or workflow.

31:25 Here’s what they value. Here’s what their workflow is. Here’s how it does. Here are the reasons why they do click this Excel spreadsheets file thing, drop down filter, cut half the records, paste it here, and make this thing. There’s a reason why they do what they do. Otherwise, they wouldn’t be doing it in the first place, right? It’s very important to understand that. It’s it’s important to understand the workflow as a whole, the values behind the workflow as a whole, and then to take a couple steps back. If you’re doing this as an organizational scale, and think of what will happen if we continue doing this little bit of rejigging here, there, here, there. It’s going to be a multiplicative effect, not only for productivity.

31:25 That’s going to happen. That’s why you’re doing it in the first place, but disruption. Right? Imagine if all the SOPs and all the processes of multiple departments and all their handoffs is changing over a six to eight to 12 month period of time, and we’re talking about institutions which have 20 years, 30 years hardened SOPs, right? This is a lot of change. So coming back to change management and the incentivizing the change management, right, and making sure that you agree and you understand and you’re aware and so on and so forth, that you actually help the people who are actually doing the job at the end of the day and walk this journey with them championing this rather than the other way around.

34:14 Patrick Coyle: It’s a constant reminder that with all these advancements we have, people are still the challenge and the mindset, and the only thing that ever actually changes is a person deciding to embrace that challenge differently. And I think there’s so much to unpack there. I would really be honored if you join me on another podcast so we can really dig down into the mindset of change and how do we get people, the human intelligence, to adopt the AI intelligence. I, it’s people like you, Dilip, that are the reason that this podcast exists now, and they’re really just exciting. I can’t wait to see what goes next. So thank you so much for joining me.

34:55 Dilip Raghunathan: Thank you, Patrick. This was a pleasure. I would. I’m honored that you would invite me. Twice, absolutely. We can dive in into other topics. We haven’t even hit governance and things like that yet, right? Of these systems, that’s golden. That’s key. So I would love to continue this conversation. And Patrick, you have a pretty big rap sheet yourself, and I see all the places. And we, it’s amazing that we’ve had all this intersection, and yet we’ve just met recently, like really met met recently, right? So I’m really glad to have have you found me out of the ether, like I said. And I’m really glad we’re having this conversation, and and I’m ecstatic to be joining this journey. And I think what you do here is extremely important, right?

34:55 You’re talking about patients. You’re talking about profitability. In a way, you cannot talk about one without the other, like you said before. And also, in this day and age, you have to talk about both simultaneously, or else one side of the equation won’t listen. And that’s exactly what I talked about, even with AI adoption, right? Which is, if you’re just doing a great technology solution, but the other people are using it or not in on board, you’re going to fail, so it’s got to be profitable, and it’s got to be for the patients, right? You just got to bring those two worlds together, and we’re going to shoot over this guy. So yeah, amazing stuff. But keep doing what you’re doing, and glad to be a part of this.

36:12 Patrick Coyle: Yeah, thank you. Patients and profitability are not competing priorities; they rise and fall together, and they’re the same mission. So it’s time we start talking about them together. Thank you so much for your time. Can’t wait till next time. Thank you.

36:24 Dilip Raghunathan: Thank you, Patrick. Appreciate it. Bye.