Latest / CXOTalk / What If AI Could SPEED UP Clinical Trials by 90%? | CXOTalk #877
Transcript
- 0:00Clinical trials are fundamental to developing new medications,
- 0:04but they face significant challenges, including lengthy
- 0:08timelines, high costs, and complex data management.
- 0:13I'm Michael Krigsman, and today on CXO Talk Episode 877, we're
- 0:19discussing the impact of AI on this critical part of our
- 0:23healthcare system. Our guest is my old friend Jay
- 0:27Farrow, Executive Vice President and Chief Information Technology
- 0:32and Product Officer at Clario, a leading provider of clinical
- 0:38trial technology solutions. I oversee IT, product, data
- 0:43security, marketing and corporate communications, all
- 0:48the things that you normal normally would find in the
- 0:51traditional CIO role, all the things that you normally would
- 0:55find in the the CTO role, enterprise architecture,
- 0:58etcetera. And then product and marketing.
- 1:00We also get to build the solutions that empower clinical
- 1:05trials all around the world. And I get to have a hand,
- 1:09although I leave it to the experts in the marketing of that
- 1:12of those products all around the world for our customers and
- 1:15ultimately the patients that we support.
- 1:17Jay, it's in a very unusual role again, Chief Information
- 1:23Technology and Product Officer. How does a role like that
- 1:28actually come together? I've I've have not heard of
- 1:31anybody with that exact title before.
- 1:33To me, it comes down to trust and delivery and transparency.
- 1:37I work for a terrific CEO. My team and I deliver.
- 1:42We're transparent and we own issues.
- 1:45We're business focused first, technology focused second.
- 1:49In no way am I minimizing the important of staying ahead of
- 1:52technology trends and leveraging emerging and transformational
- 1:56technology. That is not the point.
- 1:58But the point of an organization, especially with
- 2:00what we do is serving our customers.
- 2:03And it's not just about the technology, it's about
- 2:06empowering clinical trials. So walking a day in the life of,
- 2:10of what a patient goes through, of what a, a clinical trial
- 2:13investigator goes through a trial site, what our sponsors
- 2:17deal with when they're the ones, you know, funding a trial and
- 2:20truly understanding the business of what we do, I think has built
- 2:27trust with my CEO that I can, I could have handled A broader
- 2:31role. We have talked for years about
- 2:34understanding the language of the business, but yet I still
- 2:37see so many CIOSCTO struggling with how to do that with a lead,
- 2:41with technology and then kind of maybe put a business wrapper on
- 2:44it. And it really starts with an
- 2:47understanding of the value chain, acquiring customers,
- 2:50delivering your solution, whatever that may be, or your
- 2:53product, servicing the customers at the end, making sure that
- 2:57they are having an amazing experience.
- 2:59You're staying ahead in the innovation cycle.
- 3:01So you're, you're constantly delivering and all of that stuff
- 3:05is powered by technology, but it's not technology that's the
- 3:09end game, right? It, it's removing friction, it's
- 3:12delivering value, it's delivering services on time.
- 3:15It's making sure things are up. Of course, it's delivering high
- 3:19quality data when you say you're going to do it, it's the new
- 3:22endpoints that are being brought into a clinical trial.
- 3:26And when you can do that consistently and show that you
- 3:29have a good understanding of whatever domain you're in and
- 3:33you're having open and transparent conversations with
- 3:36your CEO about your aspirations. Part of this came from me just
- 3:40raising my hand and saying I'd like to, when you do have other
- 3:44opportunities, I'd like to be considered.
- 3:45And and so, you know, part of that has come through delivery,
- 3:50part of it has just come through transparency and good
- 3:52relationships. So what is the benefit or the
- 3:56value of combining these multiple functions under this
- 4:03particular umbrella? Well, my CEO only has one throat
- 4:06to choke. He knows where to go if
- 4:10something's not on time or it's delivered with with excessive
- 4:13bugs or defects. But all kidding aside, we own
- 4:19the solutions, the technology portions of the solutions all
- 4:23the way from inception, from a product management point of
- 4:26view, all the way to marketing. Now we have other groups that
- 4:30are responsible for delivering our services, our solutions, our
- 4:35science, all of those things. And they're amazing at what they
- 4:39do. That doesn't follow under my
- 4:40remit. But the other advantage it
- 4:43gives, Michael, is we're in a lot of different areas, right?
- 4:45We do what we deal a lot with cardiac safety, with
- 4:48respiratory, with medical imaging, what's called ECOA or a
- 4:52electronic clinical outcome assessment.
- 4:56All of those things are different, but there are common
- 4:58threads. So being able to unify those
- 5:01technology solutions or those products and those solutions
- 5:05into one tech stack and provide a more seamless experience for
- 5:10our customers. And our customers are of course
- 5:13the sponsors who are underwrite the, the clinical trial that
- 5:16could be large pharma, perhaps acro or a biotech, the clinical
- 5:20trial sites all around the world who are actively running the,
- 5:25the trials on behalf of the sponsors.
- 5:27And it of course, more most importantly are the patients,
- 5:30the, the people enrolled in a clinical trial wherever they may
- 5:33be around the world, right. And, and we want to make sure
- 5:37they have the most frictionless, seamless experience they have.
- 5:40They're all overworked. They may be going through tough
- 5:42times. If they're going through a
- 5:44clinical trial, we want to keep them front and center.
- 5:47And because this role is now unified and we have one throat
- 5:50to choke and we have a unified product management organization,
- 5:55we were able to see and leverage technology and products all
- 5:58across our ecosystem versus having siloed products or siloed
- 6:02technology. And we now talk about Clario in
- 6:07terms of an overall ecosystem in platform versus a bunch of point
- 6:12solutions. And our customers want that they
- 6:16they want a Clario to deal with, not a Clario comma XYZ solution
- 6:22ad infinitum. So the ultimate point then,
- 6:26correct me if I'm wrong, is by unifying these teams across what
- 6:34are typically silos, you're enabling a hopefully seamless
- 6:40experience for your customers. I think that's one of the
- 6:45largest benefits, Michael. I think one of the other
- 6:48benefits is internally, if you want to put kind of a, a tech
- 6:53spin on it is I'm, I'm able to, and I say I'm as a proxy for my
- 6:57team is able to see across our entire tech landscape, look for
- 7:01commonalities, look for differences.
- 7:05You don't want to build 6 of the same thing because you have
- 7:07product silos all over the organization that are kind of
- 7:11doing their own thing. O unifying that vision and being
- 7:16able to see where you can build once and deploy many is huge.
- 7:20There are the economic benefits of that.
- 7:22I build it once and I get to use it all over the organization.
- 7:25That's wonderful. Keeps my CapEx lower than it,
- 7:27you know, otherwise would be. Keeps my spend, supportability
- 7:31gets better, all of those things that we care about.
- 7:34But at the end of the day, the downstream impact is going to be
- 7:37felt by the customers, the sites and the sponsors.
- 7:40They're not getting a different experience every time they
- 7:42interact with a different product line or anything.
- 7:45They're used to seeing things a certain way.
- 7:47They're used to dealing with Clairio a certain way.
- 7:50And that is really what we're striving for.
- 7:53If you've ever walked into a clinical trial site, and I'll
- 7:55describe it for you, right? And these investigators are
- 7:59working their tails off. They're terrific, hard working,
- 8:01brilliant folks. You see stacks of devices,
- 8:07laptops, you know, respiratory devices, cardiac devices and
- 8:11stickies on each one of them to, to kind of talk about, you know,
- 8:16which company it's from, which study it's in all very, very
- 8:20important to keep everything organized.
- 8:22But I can't control what my competitors do.
- 8:26I can't control what other parts of the clinical trial ecosystem
- 8:29do. I can't control what Clario
- 8:31does. And what I really want more than
- 8:33anything is when an investigator or a patient deals with Clario,
- 8:39whether it's a BYOD app during the life of a trial, a connected
- 8:43device during the trial, a portal that a sponsor is going
- 8:47into is they have a frictionless, seamless
- 8:50experience. And if that sounds familiar, it
- 8:54should because really every industry wants that, right?
- 8:57And, and every industry I think thinks it's unique in that while
- 9:00we can't possibly think about it in consumer terms, and I would
- 9:04argue that you absolutely have to think about it in consumer
- 9:07terms. If every time I went to Bank of
- 9:09America and I had 17 different apps, well, this isn't the
- 9:13transfer app, you got to use that.
- 9:16This isn't the loan app, that's a different app.
- 9:18This isn't the check your balance app.
- 9:20You got to fire up something different.
- 9:22You'd be like, man out, I'm out of here.
- 9:25And and and. So we really want that unified
- 9:27experience while retaining speed and quality integrity of the
- 9:32data. Very, very interesting as you're
- 9:35describing the comparison of this clinical medical system.
- 9:42But from a user experience standpoint, we're all just
- 9:47people and and we want that. We want that non siloed seamless
- 9:53experience as you were describing with a bank.
- 9:55You nailed it and and that's it. That's it.
- 9:58And, and I think the the secret is you always have, we are in a
- 10:05highly regulated industry. So in no way am I saying that
- 10:08you freewheel it just to make it ease of use and that's your only
- 10:10bar. There are regulatory hurdles,
- 10:13there are scientific, there's scientific validity that you
- 10:16have to meet. And so we are always focused on
- 10:19the science. My chief medical officer, Doctor
- 10:22Todd Rudo and I have a great relationship and the the
- 10:25hundreds of scientists that we have in our organization are
- 10:29amazing partners. But within the context of
- 10:33keeping everything valid and you know, meeting our regulatory
- 10:37burden as much as I can, I want to make that a consumer like
- 10:41experience and and as friction free as I can.
- 10:45And I just want to tell people that you should ask questions.
- 10:50If you're on Twitter, use the hashtag CXO talk.
- 10:55If you're watching on LinkedIn, just pop your questions into the
- 10:59chat. And truly it's a unique
- 11:03opportunity to ask Jay Pharaoh pretty much whatever you want.
- 11:06So, so I hope you folks take advantage of it.
- 11:10So we have, let's see. Our first question comes in from
- 11:14Chris Peterson and he says this is a touchy subject in Clario's
- 11:19niche. Are you seeing or expecting
- 11:22customers hedging because of research funding cuts?
- 11:26We're certainly keeping an eye on that and you can't help but
- 11:29read everything that's going on in the news and wondering what
- 11:33the knock on effect is. And thank you, Chris, my good
- 11:36friend Chris, who has been a regular viewer of yours.
- 11:39I know that. And and I am going to say that
- 11:43the last time I was on here, which was years ago, he asked a
- 11:45question. So I am honored that he showed
- 11:48up again so I didn't scare him off, which is great.
- 11:50But yeah, we obviously have to keep an eye on that.
- 11:54It's an uncertain environment. I don't think you can pop open a
- 11:56news app and not see something about volatility, funding cuts.
- 12:04I, I personally, Jay Pharaoh's opinion, want to keep him, you
- 12:07know, the United States on the forefront of, of, of medical
- 12:11research. But there's so much good
- 12:13research going on all around the world.
- 12:17You just don't want to see any let up in, in clinical research
- 12:22going because people are depending on it.
- 12:24There are life saving, potentially life altering cures
- 12:28or treatments in, in the pipeline that we want to prove
- 12:33that are safe if we can and then get them out into the to the
- 12:38people. So we're keeping an eye on it.
- 12:43Clearly, Chris, we, we, we don't want to see that type of impact.
- 12:48The clinical trial space, there's a long tail in it.
- 12:51Generally these things are not done overnight.
- 12:57And, and hopefully I think we can withstand kind of the, the
- 13:01chaos that's happening right now when we get a little bit more
- 13:04clarity and predictability in the coming weeks and months so
- 13:07that we can get back to business.
- 13:09But right now we're we're keeping on, keeping on.
- 13:13Now is the time to subscribe to the CXO Talk newsletter, go to
- 13:17cxotalk.com, join our community, participate every week.
- 13:22And we have another question, this time from Arsalan Khan, who
- 13:27is also a very long time. He's another regular.
- 13:30He's another regular, and we're grateful for Chris and for
- 13:34Arsalan and the other folks who listen and who just ask such
- 13:37excellent questions. And Arsalan says this.
- 13:40How important is it to say yes but also say no to all of your
- 13:48teams? You're juggling multiple teams.
- 13:51And how do you decide when to say yes and when to say no, and
- 13:55when to prioritize one versus the other?
- 13:58It is something that every CXO or head of IT or anybody that is
- 14:03responsible for software, anybody in our space has to deal
- 14:09with. There is always more supply or
- 14:12more demand than supply. Always.
- 14:16I don't care what company. I have been a CXO now since 2008
- 14:21and I cannot remember one time where I ever said, yeah, I'm
- 14:25good, I got plenty and so ruthless.
- 14:28Prioritization is absolutely key and you don't want to do that on
- 14:34an island for me. What does that mean?
- 14:36It means partnership with my GMs who run our business units
- 14:40day-to-day. It means partnership with my
- 14:43CEO, naturally means partnership with my CFO.
- 14:48And he and I are absolutely joined at the hip with what we
- 14:51want to see. We want to see near term value.
- 14:54Certainly there are strategic investments.
- 14:56We're completely aligned on payback and in the making the
- 15:00decisions that have the most benefit for the company, but
- 15:04also our customers. Now there are certain things
- 15:07that you just have to kind of pay the dues every year.
- 15:10Cyber, you never want to take your foot off the gas with
- 15:12cyber. I am in an industry of trust.
- 15:17All of us, I think to some degree are in an industry of
- 15:20trust. When that trust is broken, it
- 15:21takes a while to rebuild from that.
- 15:24So cyber is always kind of pinned to the top of our
- 15:27priority list regulatory requirements, I think those
- 15:30always get, you know, pinned to the top.
- 15:32And then there are customer requests which come in and, and
- 15:35have to be juggled with those things.
- 15:37But our salon bottom line is it comes down to ruthless
- 15:41prioritization. And I try not to say no, it's
- 15:44more of a not yet or no comma. Here's what I can do.
- 15:49Can we agree on this in 2025 and then 26, we'll do a phase two.
- 15:53I really do my best to try to deliver some value, even if it's
- 15:58a no. I prefer to be like a no comma
- 16:01no. But we have this.
- 16:02This gets you 40% of the way there.
- 16:05Can we start there and then put a plan together to to pick up
- 16:09the rest of it in 20 in in 26? So it's not just a knee jerk
- 16:16reaction philosophical. Well, this is something we don't
- 16:20do. And so therefore the answer is
- 16:22no. It's there's a, there's a reason
- 16:23there's there's a. Yeah.
- 16:25You know, I will say this, if it's a customer request that
- 16:28comes in and it's just something that's not in our wheelhouse or
- 16:32you know, something that's not a core competency or yeah, yeah,
- 16:35there are going to be times where you say, look, it's just
- 16:37you, you have to be mature and self select out of it.
- 16:41Or if it's a capability that just is, doesn't make sense for
- 16:44us to build, we are always going to try to find a way to partner
- 16:48or build or, or serve our customers.
- 16:50If it's internal, nobody wants the House of no.
- 16:56I mean, Michael, you've been at this for a minute.
- 16:58You and I probably sat here 15 years ago and talked about the
- 17:02House of no and the CIO being the CI no and all of this other
- 17:07kind of thing. Nobody wants that.
- 17:10And I try never to say that, but it's, it's always, you know,
- 17:13when I do have to decline or just say, you know, we just
- 17:17cannot get to that this year or this cycle.
- 17:21It's, it's usually followed by a comma.
- 17:23But here's what we can do. And let's figure out a way
- 17:26together, GM or CXO, how we can tell the story.
- 17:31So maybe we can get some additional funding to do this.
- 17:33So I, I, you know, there's, there's some to me, it's about
- 17:39relationships and building and being transparent about how
- 17:42decisions are being made in the company and, and why we're
- 17:46choosing to do what we're doing. OK, let's go to some additional
- 17:51questions. Joseph Puglisi on Twitter.
- 17:56Ah, Joe up in Philly. Philly Joe.
- 17:58OK, so you so friends. I know, Joe.
- 18:01Great. And Joe says, will AI or other
- 18:04innovations lower the cost of healthcare?
- 18:07This is so needed in today's economic environment.
- 18:10I think it has the potential to to perhaps do that.
- 18:13Certainly we don't look at it that way.
- 18:18We, we look at it from a quality, efficiency and privacy
- 18:25point of view. So I, I think you're foolish not
- 18:29to look at AI and say there's not going to be some sort of
- 18:32art, you know, cost arbitrage. We hear, we read about it every
- 18:36single day. I think Bill Gates said
- 18:39something the other day, Michael, that he thought it
- 18:41would replace physicians or replace many doctors or
- 18:44healthcare roles. You know, I don't know how
- 18:47comfortable we are with that just yet.
- 18:48I, I think we're at the beginning of our AIAI journey.
- 18:53To me, we still want humans in the loop.
- 18:56Given, given the regulatory hurdles and the importance of
- 19:00trust in the drug development process, humans are still very,
- 19:04very much in the loop. And where AI plays a huge role
- 19:09is tackling things that maybe machines are better at.
- 19:12Looking at large volumes of data, finding anomalies,
- 19:15assisting with document, you know, interpretation or image
- 19:20interpretation, excuse me, being able to make predictions, those
- 19:25types of things. But we're still at a point where
- 19:28I want a physician or I want an expert overseeing that work.
- 19:34But yeah, I mean, the short answer to Joe's question is, I
- 19:38mean, if there there's got to be a cosplay down the road, I mean,
- 19:42AI isn't free and, and, but to me, it's less about removing
- 19:48people. That's not the point.
- 19:49To me. It's about getting our
- 19:53scientists and our amazing people at Clairio to focus on a
- 19:56higher level of challenge and augment them with the
- 20:01technology. On LinkedIn, Greg Walter says
- 20:06when implementing AI in trials, is each application of AI unique
- 20:13to that specific trial or part of a standard AI process?
- 20:20And he also asks is the AI home grown or partner?
- 20:26The answer is yes. The AIS are models that are it.
- 20:30Well, it depends. They are limited in
- 20:32functionality to a certain a certain function and that model
- 20:37can then be used across trials. They are trained more often than
- 20:42not, they're closed models. So they are not continually
- 20:45learning. So they are closed as of a
- 20:47certain point. So whether that's a data privacy
- 20:49model that anonymizes medical images, it's a QC or quality
- 20:54control assist, a model that looks at early identification of
- 21:00quality concerns to help us ensure timelines.
- 21:04Maybe a Reed assist. We're having a, a, a radiologist
- 21:08look at images and MRIACT scan those types of things and you're
- 21:12identifying it's, it's identifying different things for
- 21:15you and it can reduce inter and intra observer variability.
- 21:22Those models are built and are fit for purpose for that
- 21:24function, but they're utilized across across studies.
- 21:29So and there's many more. We at this point we have 50
- 21:33models, 50 plus models that are proprietary home grown across
- 21:40ecoa, cardiac imaging, respiratory and precision
- 21:43motion. We do have a proprietary Gen.
- 21:48AI platform that we have built internally on enterprise class.
- 21:54You can think open AI, you know, Gemini etcetera that are
- 21:59private. We named her Claire.
- 22:03We're not super creative, Michael.
- 22:04So Clario, declare. I know that's a.
- 22:07You can see the leap we took there.
- 22:11And well, hey, you're in a conservative business, right?
- 22:13The health. Care and if you saw the AI image
- 22:17and I'll share it with you sometime.
- 22:19If you saw the AI image that AI created, we said what would
- 22:24Claire look like It it is absolutely terrifying.
- 22:27It it it is terrifying. But what I love about it is a
- 22:31year ago within the organization, if you had asked,
- 22:36I don't know 1000 people, what is Claire?
- 22:39They, they wouldn't have known maybe a year and a half ago
- 22:42today you're now hearing people say, well, we're going to put
- 22:47that in Claire or we're clearing it.
- 22:48And, and it's become part of our lexicon because we've built
- 22:52capabilities to do a protocol summary, user story writer, RFP
- 22:58and contract analysis, translations, you know, document
- 23:03chat where you have a large volume of documents and you're
- 23:06trying to ascertain what exactly is and so many more.
- 23:11And the line of people behind our AI team with just use cases
- 23:15that for, for Claire and what we're doing is huge.
- 23:19But going back to the question, we try to leverage third party
- 23:23wherever we can as long as it's safe, it's private, but a lot of
- 23:28our other proprietary models are built on our home grown Aquarius
- 23:32engine. When it comes to Gen.
- 23:34AI and summaries, how do you ensure that there are no
- 23:41hallucinations? For example, summarizing a
- 23:46patient record, the AI could potentially think to itself,
- 23:52think whatever that, whatever that actually means to the AI.
- 23:55But the AI says given all of these symptoms, the patient must
- 24:00also have XY or Z condition and then inserts it.
- 24:04How do you how do you deal with that?
- 24:06Constant oversight, human oversight.
- 24:10First of all, we don't deal with anything that can be attributed
- 24:13back to a, you know, a Michael Krigsman or a Jay Pharaoh or
- 24:15anything like that. We deal with pseudonymized in
- 24:18anonymized data. So first and foremost, you know,
- 24:21we're not dealing with Joe Smith or, or, you know, Jason James or
- 24:25anything like that. We're dealing with pseudonymized
- 24:29and derived data, which is, which is super important #2
- 24:35constant refinement of the model.
- 24:37We have an internal AI team under my chief AI officer who
- 24:40reports to me that I hired and brought on to the organization
- 24:44last year, Marco Tapolovich. We are ruthlessly challenging
- 24:50our models to look for bias and hallucinations.
- 24:53One of the ways that I think I'm really proud of is right out of
- 24:57the gate when when I joined the organization and I realized that
- 25:00AI was a big part of our story and we got a little bit of a
- 25:04running start. When I I think when I joined, we
- 25:06had maybe one model in production that was
- 25:09scientifically validating. Today we're at 50 and and
- 25:13growing, which is super exciting.
- 25:15But one of the ways you do that is transparency within the
- 25:18organization. You want organizational literacy
- 25:22on what AI is, what it isn't, what we're doing, why we're
- 25:25doing it, more importantly, what we're not doing.
- 25:28And so we created a group right out of the gate, including my
- 25:32General counsel, Lauren Nishtal, my Chief Medical officer, Todd
- 25:35Rudo, my CISO Mortazanisar, my course, my Chief AI officer,
- 25:40Marco that I talked about, my head of quality, Todd, and, and,
- 25:45and all unified with our passion for the technology, but all
- 25:50coming at it from a different point of view and making sure
- 25:56that everything we're doing is built on trust, transparency,
- 26:00responsibility, bias control, et cetera.
- 26:04And I urge anybody out there that is just at the beginning of
- 26:06their AI program, do not approach it as a CIO only or a
- 26:11CTO only. Don't do it, OK?
- 26:13Don't throw all your stuff in an LLM and be like, we're on AAI,
- 26:18man, we're great. You know, when cloud came out,
- 26:20you just started throwing stuff in the cloud and you wondered
- 26:23why, you know, proprietary information was suddenly gone.
- 26:26You don't want to do that. Do it in a controlled way.
- 26:29And people immediately hear that, Michael, and what do they
- 26:31say? Oh my God, I involve attorneys
- 26:34and whoo, that's just going to slow down in absolutely not.
- 26:39If anything, it has speed it up because there's no daylight
- 26:42between US and because our customers know that we're
- 26:46approaching it for a from a responsible point of view.
- 26:49And the last thing I'm going to do is put a model into
- 26:52production that my general counsel, my head of privacy, my
- 26:56head of quality, my Chief medical officer, my CSO, my CEO
- 27:02for that matter of course that that we all haven't seen and are
- 27:05are signed off on that we did it the right way and that we did it
- 27:08responsibly. So, you know, if you're tackling
- 27:12this as an IT problem, I just, I encourage listeners to, you
- 27:16know, broaden open the aperture a little bit.
- 27:19We have another interesting question again from Arsalan Khan
- 27:23on Twitter. How do you use AI on clinical
- 27:28trials? And of course, what's the
- 27:30impact? But he's also wondering, would
- 27:32removing AI guardrails and regulations make clinical trial
- 27:38trials better or faster? You remove regulation, would
- 27:44they make them faster? Maybe.
- 27:46Would they make them better? No.
- 27:48I, I, I would argue that the FDA and I have had numerous
- 27:52conversations as recently as this week with our colleagues
- 27:55over at the FDA. They are very bullish on the
- 27:57technology. They're doing it the right way.
- 28:01I think they're moving at a speed.
- 28:02I think people hear FDA and what do they think the red tape
- 28:05right, you know, grind to a halt, see you in three years.
- 28:08And, and that is just not true. I, I might have said that coming
- 28:13into the organization five years ago, it would, it would have
- 28:16been an ignorant judgement. And I have learned that our, our
- 28:21colleagues in the FDA absolutely want to do what is best for
- 28:28people. They want to protect people, but
- 28:30they want these, these drugs and these therapies to be trusted,
- 28:36safe, effective, and they want them to improve lives.
- 28:40That's what they want. And, and in so much as AI can
- 28:45impact that, they, they want to lean into AI.
- 28:50Now, the first part of this question, which I think is
- 28:53really, really good is where can AI, where does AI fit into the
- 28:59clinical trial? Now, I'll, I'll give some
- 29:01examples where I, where I'm seeing really good effectiveness
- 29:04in the trial process. We don't play in all of these
- 29:07roles and I've talked a little bit about how we use it, but
- 29:11patient recruitment and retention, finding and enrolling
- 29:16the right patients in a trial is not easy.
- 29:19There's not a line of people hands, you know, standing right
- 29:22now, you know it. It can be time consuming,
- 29:25expensive, difficult patient retention.
- 29:29You're taking a a treatment or a pill or an injection or whatever
- 29:34over time that may or may not have side effects that you may
- 29:38not have just real world things that impact you.
- 29:42Your ability to participate in that trial and retaining
- 29:46patients is key and it's an important challenge.
- 29:48And so minimizing dropouts is absolutely key.
- 29:54What can AI do about that? Well, first of all, it can.
- 29:57It can manage or analyze large data sets, potentially
- 30:02electronic health records, genetics, social determinants.
- 30:06It can identify potential candidates who are likely to
- 30:09meet trial criteria faster, more effectively using predictive
- 30:16capabilities. It could predict drop out risks
- 30:20so that those can be mitigated ahead of time where you can say,
- 30:24hey, look, there's a a chance that Michael may or may not make
- 30:27the cut because of we're seeing those behaviors.
- 30:30And that way I can intervene potentially and keep Michael in
- 30:32the trial. That's a huge application in
- 30:35ripe for AI disruption. I think trial design and
- 30:39optimization, where we do play a role.
- 30:43Trials are complex, intentionally so, because
- 30:46they're important and there's a lot of science behind them, and
- 30:50you want scientific rigor, you want meaningful results.
- 30:54But a protocol can be very, very complex.
- 30:57You're balancing patient safety, statistics, cost effectiveness,
- 31:05so many other things. And AI can simulate different
- 31:11trial designs and predict outcomes with various
- 31:13configurations. Now you wouldn't want to rely on
- 31:16that alone, but it can at least directionally move you in a in a
- 31:22path to optimize a protocol and looking for the most effective
- 31:26study designs, not just cost effective overall effectiveness,
- 31:32which ultimately would have the benefit of speeding up the
- 31:34development process, which is is exciting.
- 31:36And those are just a couple of top of mind examples where I, I
- 31:40see AI making a, you know, a huge, a huge difference.
- 31:45We have another related question that's come in from LinkedIn
- 31:49from Lori Noregov. Before we get to Lori's
- 31:54question, I just want to remind everybody that we have shows
- 32:00like this every week. Subscribe to the CXO Talk
- 32:04newsletter. Just go to cxotalk.com so you
- 32:06can sign up. We want you as part of our
- 32:10community so that you can participate and it's fun and
- 32:16it's great. So sign up for our newsletter.
- 32:19And here is Lori's question. He says, Does regulatory
- 32:26compliance create limitations for some of the more advanced AI
- 32:31use cases? If yes, what would those be?
- 32:37And how do they work with regulators for making innovation
- 32:42more accessible? The regulatory environment being
- 32:45what it is, and keep in mind this is not the FDA only.
- 32:48You have the EMA and you have other regulatory bodies all
- 32:51around the world all with their own criteria.
- 32:53The FDA is very thoughtful and very measured in what they do,
- 32:59as are many other regulatory bodies filled with smart people
- 33:02who want to do the best for the most.
- 33:06This is to me about partnership with the regulatory bodies and I
- 33:12don't mean like Co development or anything like that, but
- 33:17understanding what the regulations are, understanding
- 33:20the hurdles and the barriers that we can that, that we can
- 33:23overcome. Showing them.
- 33:26I'll, I'll be honest in the question is such a good one.
- 33:28And I think, and I don't want to come off as an FDA Homer or
- 33:33anything like that, but as I've gotten to know them, I, I will
- 33:35say they want to do the right thing.
- 33:39And I don't see them at all as a quote obstacle or anything.
- 33:44I think a bigger obstacle is organizational readiness.
- 33:47If you said Jay is at the FDA or regulatory body and to be sure
- 33:51we are in a really highly regulated and for good reason.
- 33:55We're in a highly regulated industry and we have to be
- 33:58right. Patient safety is absolutely
- 34:01paramount and trust is absolutely paramount and the
- 34:07science behind everything we do is absolutely important.
- 34:11The reason I say that it's organizational readiness is
- 34:14because you're dealing with very complex.
- 34:18You're dealing with an industry that does not pivot overnight.
- 34:21We've seen it with healthcare over time.
- 34:24We saw it with fintech in the 90s and the 2000s, etcetera.
- 34:29If you had gone back in time and asked somebody in the 90s, hey,
- 34:31how safer is it to to use a cell phone for millions of dollars of
- 34:36financial transactions, they would have looked at you like
- 34:39you sprouted and tell that'll never happen.
- 34:41You got to have people, you got to have whatever.
- 34:43And here we are today moving money all around the world at a
- 34:46moment's notice. It's going to change, but you're
- 34:49still seeing organizational resistance.
- 34:52I, I think in some ways for good reason, because there's a, a bar
- 34:56of rigor that you want to hit. Part of it is just culture
- 34:59change and making sure that people understand the power of
- 35:03AI in this truly unique fundamental force that we have.
- 35:07So a lot of it comes from education as well in overcoming
- 35:12some of the organizational inertia in older thinking.
- 35:16I, I mean, look, I'm, I'm not a want to be too hard on people,
- 35:18Michael, but I mean, if you had walked into A and, and, and
- 35:21we're hanging out with a 25 year veteran in the fintech space in
- 35:251994 and said, I'm going to paint a picture of what the
- 35:29future is going to look like, they would have been like, no,
- 35:34let me tell you the 52 reasons why that's not going to work.
- 35:37Lori clarifies his question to say How do you work with
- 35:41regulators to make innovation more accessible?
- 35:44I was at a conference not too long ago, Lori, you, you'll
- 35:46appreciate this. And it was an FDA panel and, and
- 35:56one of the FDA guys that, and I won't say his name because I, I,
- 36:00I just, it's not important. But he came off and I, I greeted
- 36:05him because I just was blown away by the, the discussion.
- 36:07And it was all about AI and what they thought of it.
- 36:10And it just was in, it went in a totally different direction.
- 36:13And I said, and I almost asked Lori's question verbatim.
- 36:17And I was like, how do we partner with you?
- 36:20You can e-mail me. And I was like, oh, oh, well,
- 36:24that's, I mean, that's interesting.
- 36:28You know, we, we want to hear, we, we want to learn.
- 36:34But to put a finer point on, I would say proactive
- 36:37collaboration, engage early. How do you do that?
- 36:39Well, talk to them, OK, this isn't an ivory tower that I, I
- 36:46will say every FDA person or every regulatory person that
- 36:48I've ever talked to has been willing to have a conversation.
- 36:51So establishing clear communication and being
- 36:54proactive with the regulatory authority, whether it's the FDA
- 36:58and, you know, EMA, etcetera #2 understand regulatory
- 37:04frameworks, show that you're compliant with existing
- 37:08guidelines, making sure that you're familiar and you
- 37:10understand what they're trying to accomplish and that you're
- 37:13not coming at it from a purely tech point of view.
- 37:15And be like, yeah, but if we could just release the
- 37:18guardrails and let this, you know, thing go nuts, imagine all
- 37:21the good probably, yeah, maybe some truth to that.
- 37:26But I think the reality is that the regulatory frameworks exist
- 37:31for a reason. They change over time.
- 37:34Guidance has changed over time. If you had said 20 years ago,
- 37:38we're going to have an AI reasonable use policy, I think
- 37:41most companies would've been like, why the hell do I need an
- 37:44AI reasonable use policy, right? So making sure you're familiar
- 37:48with existing guidelines and, and that you're adapting your AI
- 37:54to fit that compliance, I think that starts you off in a, in a
- 37:58very good conversation with, with regulators, making sure
- 38:01that you emphasize safety and risk management with regulators.
- 38:06The FDA often requires risk management strategies.
- 38:09They don't want to hear about the art of the possible without
- 38:11the other side of the coin. What are you doing to manage
- 38:14risk to ensure that the AI technologies that you're
- 38:18proposing are safe and that any potential risk to perhaps the
- 38:25patient, the study, etcetera are are are mitigated?
- 38:29And the last thing I'll say, the last one is develop a regulatory
- 38:34friendly development practice, which I kind of harped on at the
- 38:38beginning, following best practices, having SO PS in place
- 38:45that can pass regulatory scrutiny, transparent and
- 38:49training models, bias mitigation, all of those types
- 38:55of things. Build trust with regulators.
- 38:57And on this topic of transparency, Arsalan Khan from
- 39:02Twitter comes back and he says if we want transparency across
- 39:06industry, should companies share their next AI with other so
- 39:12collectively whole industries can get better?
- 39:15And what about the role of gatekeepers in all of this?
- 39:20I do think there's an opportunity across industries to
- 39:26promote transparency and Dr. positive outcomes across the
- 39:31industry, whether it's healthcare, life sciences,
- 39:35fintech, etcetera. I think I, I think there are
- 39:40some benefits of doing that. Certainly speed collaboration
- 39:45for innovation and you're going to get more faster generally
- 39:49public confidence. I think when they see
- 39:51collaboration and trust and transparency across industry,
- 39:58there's certainly an opportunity for better regulation.
- 40:01And you're seeing that already with Europe, right?
- 40:06Their AI, their AI regulations are not just for life sciences
- 40:11or healthcare. It's AI in general matter of
- 40:14time before the United States does it either at the state
- 40:16level or the OR the federal government level.
- 40:18Certainly you have some draft guidance already out there, but.
- 40:26I think it speeds that up and I think you get a higher quality
- 40:30of regulation when there's more collaboration and sharing.
- 40:32So there are a lot of reasons to to do it.
- 40:36I'll just invention for folks that are interested in Europe,
- 40:41European adoption of AI and regulatory efforts.
- 40:46Just a few weeks ago, we had two members of the House of Lords
- 40:50discussing these exact issues here on CXO Talk.
- 40:55These are folks who are creating policy in the UK.
- 40:59So if you're interested in this topic in Europe and the UK
- 41:02specifically, just go back and look at CXO talk we had.
- 41:06It was a really great discussion.
- 41:08Well, I'm sure, I mean, look, they are always at the forefront
- 41:11of regulation. And I don't mean that in a
- 41:14snarky way, but I mean I I think they've done a nice job in
- 41:18balancing innovation with risk mitigation, transparency and
- 41:22accountability. It's not perfect, nothing ever
- 41:26is, but the E the EUAI Act, I mean, they came out of the, I
- 41:35mean, what is that in 21? I think they launched that and
- 41:38and which is crazy. It it, it, it was very, very
- 41:44well thought out. And what I like about it is it
- 41:47triages risk, unacceptable, high, limited, minimal,
- 41:54etcetera, right. So this, this risk approach
- 41:58where, hey, if it's the system that poses little to no risk,
- 42:03it's an AI in a video game, it's an AI in a spam filter.
- 42:08Have at it. You know, you're going to have
- 42:10to check a few boxes, you're going to have to do a few
- 42:12things. But yeah, we, we don't want to
- 42:15slow stuff like that down, right?
- 42:17And and it it seems to me they at least tried to right size the
- 42:22regulation which is which is good.
- 42:24So we have another question from Greg Walters, Who is aware that
- 42:31you have a background in cement. You can tell us about that and.
- 42:36Outing me like that, Greg. All right.
- 42:38And Greg Walters wants to you to compare AI to quote cementitious
- 42:46materials, given the fact that concrete has been around for
- 42:51thousands of years. Every industry is more alike
- 42:55than not. I, I know we like to not think
- 42:58that, right. I, I, I think every industry
- 43:01thinks it's a, a perfect little snowflake, Michael, that you
- 43:06know, that that is, it's a Unicorn, every little thing.
- 43:09And to be sure, every industry has nuance and has its own
- 43:13lingo. It's got its own ecosystem and
- 43:15it's got its own special spin on a regulation or whatever.
- 43:21But I can tell you the discipline of product
- 43:24management, technology management, financial
- 43:27management, software development, security, etcetera
- 43:34is 80% the same across industry. So you if you look at
- 43:37manufacturing my time as the CIO of the Quikrete companies and
- 43:40even before that, you know at AIG or the American Cancer
- 43:44Society or EarthLink, you were doing many of the same things.
- 43:49Now the size, the scale, the severity may be different, but
- 43:56when are you not protecting customer data, when are you not
- 44:01trying to deliver high quality results on time?
- 44:04When are you not trying to leverage emerging technologies
- 44:08to deliver a higher quality product or have a, an amazing
- 44:12customer experience or improve operations or drive revenue.
- 44:18And, and I would challenge CI OS and CT OS and aspiring CX OS to
- 44:24think about things that way and not box yourself into one
- 44:27particular industry. But how is cement like AII know
- 44:32there's a punchline, Greg, somewhere out there and I don't
- 44:34know what it is. Maybe, maybe somebody will
- 44:38suggest it, but. Joe P on Twitter comes back and
- 44:42he says, how do you keep your team focused on the outcomes and
- 44:45avoid becoming enamored with the technology?
- 44:49In other words, avoiding shiny new object syndrome.
- 44:53Education transparency, I'm building financial literacy
- 44:57education, educating the team on why we exist as a company and
- 45:01what what does that really mean? That those are all a lot of
- 45:03really flowery words. Site visits, discussing with
- 45:08patients, showing them what we do, visits from, you know,
- 45:12different departments, building those relationships so that
- 45:16people, no matter where they are in the organization, in my team,
- 45:19I don't care if your health test level 1, I don't care if you're
- 45:22a coder and you're a junior person that just joined the
- 45:24team. I want them feeling the
- 45:26importance of what we do. I want them recognizing that
- 45:30there's a patient at the other end that is using an app on
- 45:32either on a provision device or a phone or is using a connected
- 45:36device, whether that's a a blood pressure device or a respiratory
- 45:39device in clinic that we built or that we built an interface
- 45:44for or that we have AI integrated with.
- 45:46I want them understanding that what they're doing has real
- 45:50human impact around the world. And when you do that, yeah,
- 45:55you're still going to get excited about the technology.
- 45:57It's what we do, but it's technology with purpose.
- 46:01It's not technology for technology.
- 46:03So we try to do a lot of education, Michael and Joe P,
- 46:07but and keep people focused. So what does that mean in
- 46:11practical terms? Site visits, site education,
- 46:14voice of the customer type activities, making sure we're
- 46:17hearing from patients, making sure we're showing teams the
- 46:20impact of the technology that they're building and that
- 46:22they're delivering, whether that means, you know, hearing
- 46:26directly from customers, etcetera, so that they
- 46:28understand the impact. It wasn't just a bunch of code
- 46:31on time, which is nice. I want that, but there's purpose
- 46:35behind it. On this topic, Arsalan comes
- 46:38back again, and he's wondering about the future of AI.
- 46:44Let's keep it focused on clinical trials.
- 46:47And, he says, as AI becomes increasingly more of a
- 46:52commodity. If DeepSeek has showed us
- 46:54anything, it's that I, I think the idea of a, if a trillion
- 46:58dollar, one-size-fits-all one magical product to rule them all
- 47:03is not going to be long for the world.
- 47:06I, I, I think there will be some commoditization of, of, of
- 47:10technology over time. And I think what you'll see is
- 47:15far less one giant model and far more fit for purpose.
- 47:21What, what the future I see, I see some amazing things
- 47:24happening in the future, particularly in healthcare, in
- 47:27clinical trials, keeping it speed, improving in quality
- 47:33patient experience. What, what does that mean?
- 47:38It means if a patient is having a bad day or there's, you know,
- 47:42there's, there's just you're, you're, there's just some
- 47:44empathy that even models today can show more time with an
- 47:49investigator or a physician. If it's healthcare, physicians
- 47:53will tell you constantly and nurses will tell you constantly.
- 47:55I'm sure you everybody knows this.
- 47:57They spend more time entering things into an EMR than they do
- 48:00with a patient. Emr's are designed for one
- 48:03thing. What do they do?
- 48:04Or two things to bill and to keep street legal.
- 48:07And I'm not trying to be depressing or snarky or anything
- 48:10like that. And I don't mean anything bad to
- 48:12our friends at the big EMR companies, but there's a huge
- 48:15opportunity for AI whether to listen to the physician as he's
- 48:20talking, to transcribe notes, where they can turn around and
- 48:24spend even another minute or two with a patient and improve that
- 48:27patient experience. So I am really excited about
- 48:30what the future holds, particularly for patients and
- 48:35better outcomes. The through line of our
- 48:39conversation seems to be customer experience and broaden
- 48:44it, patient experience, whatever domain that might be.
- 48:47I think that's true. I, I look, if not them, then
- 48:49who? I I mean, if you're not doing it
- 48:52for for them, then why are you doing it?
- 48:55And and look, you and I, if I take off the motherhood and
- 49:01apple pie hat and say, OK, Jay, yeah, patient experience is
- 49:04super important. But really, companies exist for
- 49:07shareholder value and to make money.
- 49:09Fine, if you're not wrong. I would argue that one of the
- 49:14best ways to do that is an amazing patient experience and
- 49:17an amazing customer experience. If I'm making my sites happy or
- 49:21at least less mad at me or am I providing A frictionless
- 49:25experience and they always know what the other end is, a company
- 49:29or a team that is working hard to do the best.
- 49:32We're not always going to be right.
- 49:33We're not always going to get it right, but we're always doing
- 49:36what's important and what's the best for them and for their
- 49:38patients. If they know that, I would argue
- 49:41that they're going to want to continue to do business with
- 49:43Clario and they're going to continue want to continue to
- 49:45partner with us. So I think the two are
- 49:48inextricably linked. I, I, I think you can have an
- 49:51amazing patient experience and amazing site experience and
- 49:54sponsor experience maintaining scientific rigor, high quality,
- 50:00privacy, speed, and all the other things that we want to do.
- 50:03And you can still make money as an organization and build
- 50:06shareholder value and do all of those things.
- 50:08And I would argue that if I can do the first one really, really
- 50:11well and I'm making good choices, the second one will
- 50:13come. What advice do you have to
- 50:17business and technology leaders in the enterprise when it comes
- 50:21to developing AI projects and scaling those projects as you
- 50:26have done? Don't do it in a silo.
- 50:28Whereas my old boss used to tell me a highly polished cylinder of
- 50:31excellence. Don't look.
- 50:34Look partner, you know. Talk to your partner ecosystem,
- 50:38other CI OS, other CX OS, talk to your suppliers, understand,
- 50:42educate yourself, but talk to your, your peers in the
- 50:45organization and work with them, not against them.
- 50:48Whatever you do, don't make it a, a, a tech only, tech only
- 50:53endeavor. In no way does that minimize the
- 50:57importance or the, the excitement about the technology.
- 50:59But I promise you, if you can do all of those things, it will
- 51:02speed up what you're trying to accomplish.
- 51:05And is always Michael, and I think I said this on our first
- 51:07interview. Think big, start small, scale
- 51:11fast. Think big, start small, scale
- 51:13fast. Great advice.
- 51:16Hey, we have one last question that's come in very quickly from
- 51:20again from Luri Adnorigov, who says, do you feel like we have
- 51:25the right people training mindset, the site patients,
- 51:29nurses, etcetera to enable and provide all of the data needed
- 51:36for the AI applications or is the technology far in front of
- 51:43actual adoption? There's a huge educational
- 51:46hurdle that we have to hit. There are people at the sites
- 51:49and the, you know, with the patients, I I think they are
- 51:53just handed a lot, right? They are handed a lot of
- 51:56different devices and it and, and I promise that it's not
- 51:59because they are not smart people or hard working people.
- 52:02It's just a lot. And if I mean, the first part of
- 52:08the question is do we have the right people?
- 52:10Yeah, I mean, the ecosystem of of investigators and sites are
- 52:14amazing and and they work their tails off and they're more than
- 52:17capable of picking this up. I think it's on us to train.
- 52:21It's on us to show them the power.
- 52:23It's on us to make it simple. I think it's going to require
- 52:26patience, Michael. I do not think we're going to be
- 52:28able to hand them a magical gokulator and say, you know, and
- 52:31have them be enamoured with the tech in, in one day.
- 52:34It's going to take some time. And so we have to be, I think a
- 52:37little patient as we kind of change the paradigm.
- 52:40But I'm I'm, I'm optimistic that, you know, we have the
- 52:44right people. We just got to meet them where
- 52:46they are. And on that note, Jay, a huge
- 52:50thank you. I'm so grateful for your taking
- 52:53the time to be here with us today.
- 52:55Really, really appreciate it. My pleasure, Michael.
- 52:58Thank you for having me. And thank you to everybody who
- 53:01watched you guys who asked such amazing questions.
- 53:05You are so smart. Now is the time to subscribe to
- 53:11the CXO Talk newsletter, go to cxotalk.com, join our community,
- 53:15participate every week, and we'll see you again next time.
- 53:20Take care everybody, have a good day.