Life SciencesSeason 1 | Ep. 231 min
Petros Pharmaceuticals’ Fady Boctor and Infocepts’ Shashank Garg on self-service platforms, patient education, human oversight of AI, and putting health data in the consumer’s hands.
Fady Boctor
President and CCO, Petros Pharmaceuticals
Shashank Garg
Co-Founder & CEO, Infocepts
Also on
Many effective prescription medications reach only a limited population, because access depends on cost, physician availability and the barriers of a prescription. Fady Boctor, president and chief commercial officer at Petros Pharmaceuticals, describes a way to widen that access: a technology platform that helps consumers decide whether a prescription-grade medication is appropriate for them, with the FDA’s stated openness to technology of this kind.
He and host Shashank Garg discuss how such a platform is built on the drug facts label, how to educate patients as the questions are asked, and where data and AI are already used across pharma, from research studies to commercial training and targeting.
The conversation also covers what goes wrong when AI is left without oversight, how to manage change in a heavily regulated industry, and why interoperable health records are putting more power in the consumer’s hands.
Five Things Worth Your 41 Minutes
Boctor says many therapeutics that have been on the market for a decade or more have reached only a limited population because they require a prescription. For the flagship asset he describes, in the erectile dysfunction market, about 25% of the 30 million men in the US are estimated to seek treatment, for reasons including cost, physician access and stigma. He says the FDA has understood the issue, and since mid-2022 has said it is opening the door to technology and assistive devices that help consumers self-select.
Boctor says the platform is built on the drug facts label, which represents millions of dollars of investment and thousands of patients tested. Algorithms based on the label drive plain-language questions about medical history and medication history, building a profile of appropriate or inappropriate use in a few minutes, at the retail pharmacy shelf. Additional AI and machine learning components help complete the picture. Garg calls this an example of democratizing data and AI at scale, with the regulator on board.
Boctor compares it with a good doctor who explains why a question is being asked, whether there could be a comorbidity, and what else could be happening. The platform explains why each question matters, offers tabs for more information, and does not assume the consumer has understood, which both educates the patient and gets an accurate disposition. He says the FDA holds the company accountable to the public health value.
Boctor lists streamlining clinical and research studies, physician profiling and anticipating consumer behavior, many of them still nascent, plus security checks that validate the consumer’s identity. In commercial teams he sees Generative AI creating varied physician personas for sales rep training, and AI for targeting geographies, prescribers and payers. Garg adds conversational BI over structured and unstructured data, including a project for a CFO’s office on clinical trial cost management.
Boctor says AI can go wrong if a company deploys it without understanding its source model and assumptions, since AI cannot identify or correct its own bias. In his example, a vendor’s AI interpreted research study results without enough oversight, was 90% or better at transcription, but missed the implication of certain words, and the FDA took notice. Garg describes governed experimentation, with explainability and compliance checks independent of the data scientist who built the model.
Boctor says, because the industry is heavily governed by the FDA, it is important not to experiment on vital consumer engagement or healthcare delivery. He recommends leveraging existing technologies that have been refined outside your industry, which also builds credibility. Garg adds that in a regulated industry, the smaller the perceived change, the higher the chance of adoption. Boctor also sees interoperability and data exchange, including TEFCA and QHIN, as a safer place to experiment.
He points to a shortage of about 120,000 physicians, with primary care hardest hit, and to consumers gaining access to consolidated health records, which together mean consumers are starting to make more of their own decisions with their own data. He expects significant leaps forward by 2025. Garg adds that the cost of training and running large language models keeps coming down, which, with the right balance of experimentation and oversight, makes the next 18 to 24 months look very different.
And now we’re starting to empower consumers to make those decisions and seek those treatments for themself, leveraging their own custom healthcare data.
There always has to be human governance, there always has to be oversight.
It’s important that we do not experiment on vital consumer engagement or vital healthcare and delivery.
The Infocepts Point of View
Two ideas run through this conversation. First, the same data that serves operational reporting can serve an entirely different ecosystem when it is paired with the right technology and a supportive regulator. Second, in healthcare the model is only as safe as the human governance around it, which means knowing its source, its assumptions and its limits.
Shashank Garg’s view is that organizations running many AI initiatives need a way to pressure test them, with experiments loaded alongside their expected business value, data quality, and explainability and compliance checks that sit independent of the data scientist who built the model.
For life sciences leaders, three questions follow: which of your assets could reach more people through a consumer-facing platform, where does your AI need a human reviewing the output, and what proven technology from outside your industry could you adapt first?
About the guest
Fady Boctor
President and CCO, Petros Pharmaceuticals
Fady Boctor is president and chief commercial officer of Petros Pharmaceuticals. He has about 25 years in the industry, with large players across respiratory, cardiovascular and hospital-grade antibiotic products, and now focuses on technology that helps consumers self-select over-the-counter access to prescription-grade medications.
Facing this problem?
Talk to our Life Sciences experts about consumer platforms, AI governance and the data foundation behind them.
#1 on Gartner Peer Insights, 3 years running · 97.2% client retention
Bring This to Your Team
A private briefing on data and AI in life sciences, covering use cases, governance and adoption, set against your own data estate.
Shashank Garg 00:07
Welcome to the Intelligent Leader Podcast. I’m your host, Shashank Garg, and today we have a very exciting leader with us. We have Fady Boctor, Fady is the President and Chief Commercial Officer of Petros Pharmaceuticals. Just to introduce Fady, he’s a bit of a data nerd, loves data and AI and infusing technology to do business. In fact, he’s, it wouldn’t be unfair to say, Fady, that you’re almost betting the firm, on this. I’ll let you talk about the recent announcements around what Petros is trying to do.
Fady Boctor 00:47
Thank you, Shashank. It’s good to be with you today and great topics, worthy of discussion. Yeah. In many respects, we have high bets, high bets on technology, the infusion and the conversions of technology led by way of artificial intelligence, but so much more. And so, yeah. Just about 25 years in the industry, been with the big players, the traditional innovative products. Everything from respiratory to cardiovascular to hospital grade, carbapenems, antibiotics. So phenomenal side and journey on the innovation part. The pharmacological innovation part. Recently, found myself gravitating towards leveraging tools of old, but in new ways for new purposes delivered in expanded access and leveraging various technologies. And I’m pointing directly around technology that helps consumers appropriately, self-select as over the counter options of prescription grade medications. So, yes, to answer your question, a bit of a data nerd, high bets, high risk, high reward, and loving the journey that I’m on.
Shashank Garg 01:51
Fady, do you wanna describe for the audiences a little bit about what exactly that means, the self-service, self-selection, and how does that change the pharma industry, why are you so excited about this and what’s the opportunity here?
Fady Boctor 02:06
So the opportunity really revolves around leveraging fantastic therapeutics that have been on market for a decade plus, but maybe have only had a certain population that have accessed it because they’re limited to prescription access alone. In our particular case, for our flagship asset that will sort of become the debut for this new technology. It’s in the erectile dysfunction marketplace. So, to quickly just quantify the answer, 25% of the 30 million men in the US have been estimated to seek treatment. That means three quarters of 30 million men in the country have not, simply because, among many reasons, cost, physician access, prescription medications, taboo stigma, embarrassment, any number of those reasons, they bind prescription medications to a limited population. So the technology we’re going after, and, I think the marketplace that’s currently before us, the FDA has understood this. It’s nuanced prescription medications that are not easily converted to over the counter, such as ibuprofen or antihistamines, but product that have significant contraindications and have significant concerns, still safe, largely safe, but with nuanced contraindications. How do you put those in the hands of the consumer without the burden of a prescription?
Fady Boctor 03:28
So the FDA has mentioned before, middle of 2022, indicated that they’re opening the door for technology assistive devices and for manufacturers to be innovative in the development of those technology platforms. We’ve been working on that since, and we currently have a proprietary web application that is in pursuit of a software as a medical device approval that essentially helps the consumer to appropriately self-select or deselect based on their clinical profile to purchase a prescription grade product over the counter. And that is the exciting frontier. And it’s critical because we think those medications, their innovation still has not been fully tapped into by the subject population, the patient population it was intended for.
Shashank Garg 04:10
And that’s exciting to hear. And Fady, what just you’re describing in my mind, I’m going through, this business is not even possible without the right use of data, web technology, and building that self-service platform. You just couldn’t do it, the way you’re doing it is the only way of doing it. And the fact that FDA has come around and said that we will support it, is, you couldn’t be in better times. This is the best time to be. And in things like these. Right?
Fady Boctor 04:43
Agreed. The landscape is ideal. In fact, 10 years ago, this idea came up with FDA and they called it back then, ensure non-prescription utilization, and the acronym goes on, but nothing happened, nothing formed from it. So, to your point today, the market landscape, the average American has spoken loud and clear. The American public has spoken loud and clear. The FDA has started to acknowledge technology’s emerging, and there is this new vacancy, this new emerging potential to meet tremendous need of self-care with prescription grade medication. So to answer your point, yes, this is the right time to do so.
Shashank Garg 05:20
Awesome. And Fady, just to introduce a little bit about myself and the firm, InfoCepts, the sponsor for this podcast. So we’ve been in the data analytics and AI industry for two decades, and what you just described is the exact set of transformations that excite us. Data can serve many purposes. It can be operational reporting, it can be keeping everybody informed. It can, ensuring that we support decisions, executive decisions at the right time. In the data business we always talk about giving the right data at the right time in the hands of the right stakeholders, and that achieves a certain outcomes. But now to be able to take the same data, marry it with the technologies, the regulation, and create an entirely different ecosystem where you can serve many more men in this case, the American population, is a huge transformation. And there are several such examples that we are supporting with many clients, we work across industry clients, retail life sciences, media, and we are starting to see many such transformations happen across the industry. But it’s really a pleasure to have you, personally, who’s driving this transformation at Petros, and to talk about that. Just moving a step forward. Fady, could you lay out for our audiences when you talk about this platform, what does it really mean? So what’s happening behind the scenes and what role data and AI is playing in implementation of such a platform?
Fady Boctor 07:04
Yeah. I appreciate. I think one key component for any technology is, what is the foundation that it’s built upon? And in our case, it’s built upon the drug facts label. So, the drug facts label represents millions of dollars of investment, thousands of patients tested in consumer, sort of, examined. And we understand what is safe use and what is not safe use. So, we take the drug facts label, we build algorithms based on that drug facts label, those algorithms then fuel certain questions, consumer layman friendly questions. And these questions, basically ask the consumer about their medical history, their medication history. The algorithm of the DFL or the DFL is a source driving the algorithm behind each of the questions, helps the consumer understand, is this appropriate for me or isn’t it? So as they’re answering questions, they’re establishing a profile of appropriate user or inappropriate user. And as in its most basic form, that’s what this proprietary web application is intended to. There are additional artificial intelligence components, machine learning components, that help seal the deal, help seal the picture. But that is in its essence what we’re doing. It’s supposed to be transient, just a few minutes of engagement right there in our retail pharmacy shelf and aisle. You’re engaging with this web app, within minutes you’ve been deemed appropriate or inappropriate based on a clinically defended, algorithmically driven medical history questionnaire.
Shashank Garg 08:36
That’s great. And for anyone who’s listening this, we talk about democratizing data and AI to a larger population within the organization. What we’re talking about from hearing from Fady is an example where they are democratizing this information and determining eligibility for a particular drug which is approved, clinically approved by the FDA. So ensuring that the regulator is fully on board and doing that in a safe way. And the fact that we can do that at scale, is just fantastic and the right time to be in the data and AI industry. So thanks for sharing that, Fady. Moving on and just sort of looking at the broad pharmaceuticals industry, and pharma industry talks about building such, and if I may use the generic term, patient education platforms and supporting self-care, not just to do what you are doing, which is taking prescription drugs over the counter now, but just supporting themselves. Hey, could you share how to do this right? And what can go wrong in this because it’s very sensitive. So how do we do this, right?
Fady Boctor 09:51
I think we do this right by educating the patient within the engagement. So sometimes you go to your doctor’s office and you can tell, a good doctor will not only ask you questions and write down your answers and include your answers and then make a decision, but a good doctor will counsel you and he’ll educate you or she’ll educate you. So as they go through, they ask a question, they’ll say, “Here’s why I’m asking the question, or do you understand why I’ve asked this question?” So the right way this is done is to help a deeper engagement with the healthcare environment, meaning we know that the patient is interested in receiving therapy, we’re gonna ask them questions about their health, but then do we educate them as to why we’re asking those questions? What else could be happening? What is the source of ailment? Is this a comorbidity to our larger deal?
Fady Boctor 10:35
Is there something larger underlying happening? Or, basically said, if we’re asking a question about how often you have so-and-so symptom, we’re explaining why that’s of concern to us. We do that. And not only are we correctly and appropriately educating the patient, but we’re also getting their accurate disposition. And I think that is the right way to do it. And that’s what our technology intends to do. We educate them as we ask the questions. We offer them tabs for more information. We don’t take it for granted that they’ve understood everything that we’ve said or everything that we’re asking them, which either way, the FDA holds us accountable to such a value. We couldn’t just say, we wanna take this prescription drug and convert it to OTC without a public health value. So the public health value is intrinsic in our altruistic mission. FDA holds us accountable to that as well.
Shashank Garg 11:21
Thank you, Fady. Let’s, moving on. I’m just, maybe spend some time broadly talk about where you are seeing similar use of data and AI in pharmaceutical industry in general. Where do you think this is becoming very real, and very useful? If you wanna share from your experiences and examples there.
Fady Boctor 11:43
It’s interesting because there are a diverse… There’s a plethora of diverse uses of various technologies. Everything from clinical study or research study, implementation of artificial intelligence technology to help streamline the process, to help simplify and streamline and expedite results, expedite protocol. So there’s that facet, I see that happening quite a bit. Some for good, some not quite yet ready for showtime or primetime. And there’s also utilization, to your point, in terms of physician history and physician profiling and anticipating consumer behavior. There’s a lot of technology around that, whether it’s prescription data or generative AI data that looks at, if a patient comes from X geography with X profile, here’s what they’re likely going to prefer. We see those types of models occurring. I think many of them are still largely nascent. Then there’s, obviously technology and AI that’s leveraged, in our case, which I have not to be honest with you seen in this particular field.
Fady Boctor 12:47
And that is how do we validate the consumer? How do we make sure they’re not spoofing? How do we make sure that they’re not abusing the system, they’re not counterfeiting their profile? So there’s some security elements to make sure the person who’s answering the questions is in fact who they say they are by way of age, gender, and government ID, and so forth. And so from all the facets, I would say a lot of activity occur across the board, but there’s much work to be done to validate it and to get FDA buy-in across all fronts.
Shashank Garg 13:17
I remember, so one of our clients at a large pharma firm, our teams, and you talked about generative AI as well. So just for our audiences benefit. We don’t always have to look for transformational examples. They are the ones that require a lot of buy-in right from the top. But in my mind, in the whole field of data analytics innovative AI is starting to play a very important role, something as simple as, what we talk about, conversational BI or being able to ask the right questions. We have a solution called Decision360 and just organizational information. Sometimes it’s just very hard to find data, and especially if it’s a combination of structured and unstructured data. There’s technology available where on the top of your structured data, data warehouse, data lakes, you can marry it with unstructured data and make it really easy for your operational users to just ask questions.
Shashank Garg 14:16
And they can come in a narrative, where you check for hallucinations and you’ve done it in a secure way. So at least within the walls of your organization you’ve made it very easy for people to get to answers, which even after three decades of work, that’s how long I have been in this field, is unfortunately not that easy. One of the projects we did is for the CFO’s office, around the whole topic of clinical trials, cost management, it’s huge cost involved. And the CFO’s office will, for somebody in commercial, will come down very hard and heavy and say, “Why are we burning so much money?” And just make it easy for their analysts to talk to the financial data. Just made a huge difference there. Fady, I also see a lot of excitement and possibilities around enabling the commercial organization, the ones that the sales representatives who, actually call on the physicians. Are you starting to see any interesting examples of use of AI there?
Fady Boctor 15:25
So it’s interesting you had indicated, for training, certainly, so to develop multiple and diverse perspectives in training. So reps in training with physician profiles, those physicians that like to put up smoke screens. How do you leverage artificial intelligence, generative AI to ensure that the patient, the consumer, the representative rather, is diverse enough in their ability and capacity to communicate essential elements through the various profiles. So to be able to establish a variety of personalities for the representative to train on is fantastic. And I’ve seen that starting to emerge more and more. And I think that’s really, a largely, where there is a great deal of momentum. Targeting is another area. How to target certain geographies, certain prescribers, prescriber specialties, and then the story goes on. I think targeting payers is another area where you want to understand based on history and based on behaviors and contracts, where are the best, most productive contracts. So those types of things are currently in play from a commercial perspective.
Shashank Garg 16:31
Yeah. I’ve heard of a few things that you talked about, but what’s interesting is you earlier spoke about training, which is an angle that, I personally, haven’t seen as much, but that’s a very good one. And that’s where the generative powers of AI can be leveraged. You’re essentially saying that to train a sales rep to be able to respond to or deal with all the different smoke screens that a physician is likely to put up, can be very useful. And instead of just enabling them with data, “Okay, here’s the physician order history and here’s what you’ve done with them and here’s what they like you to do.” You also use some generative powers to, make them talk to an agent, for example, before they actually show up to a physician.
Fady Boctor 17:18
That’s right. It’s, you’re giving life to a profile and then giving the ability to also be, in many respects, unpredictable.
Shashank Garg 17:26
Yeah. Talking about algorithms, Fady, I know one of the last times we were talking about you had shared that sometimes it is not as easy to get the work right. And, can you throw some light there with your personal experiences, any areas that went bad didn’t pan out the way you expected?
Fady Boctor 17:51
Yeah. I think, and correct me if this is not the scenario you had indicated, but I think artificial intelligence can go wrong if any company decides to implement it without fully understanding it’s source model. What are the assumptions? What’s the model? What’s the model that fuels it? Artificial intelligence in and of itself is not capable of identifying its bias. It’s not able to correcting its bias or correcting its model. It’s subject to whatever model is fundamental to its thinking. And so in our case we, this was a vendor that leveraged AI in their, in one of our research study results. And unfortunately they leveraged AI in a way that was not overseen well, and therefore it was, the AI was left to interpret the results. Although it did relatively well, relatively in the 90% or better, probably in capturing certain transcriptions, it missed certain key elements. It was ill-equipped to understand the difference between one word and its implications versus another. So although it didn’t necessarily do it wrong, it acknowledged the right word in the right way, but did not catch the implication of that word in this context. All that to say, yeah, FDA took notice.
Fady Boctor 19:08
And even if you’re 1%, 3%, 4% off, they’re going to take notice and that can’t happen. So I say that to say, there always has to be human governance, there always has to be oversight. And the person who established the model, the person, the vendor that that designed it should quickly and upfront identify, be thoughtful, be careful, look for these elements, pressure test and make sure that you know you’re not taking for granted. This is 100% accurate. So yeah, for all those reasons, I would say be mindful. Do not give artificial intelligence its own legs too soon. Without or without ongoing human governance.
Shashank Garg 19:43
You bring up a good point. And I’m starting to see across our clients and I’m realizing is that at the leadership level, because there’s so many AI initiatives and experimentation going on, because you want to promote experimentation, that’s how we move forward. I see some of some of the organization getting overwhelmed by just sheer number of initiatives that are going on. And to be able to pressure test everything is a challenge. So, what we’re starting to do is at least internally at Infocepts, we’ve developed our own sort of experimentation platform. We sort of call it, you know, sort of governed experimentation, right? So, you can load up all your use cases and the anticipated business value and the checks and balances that you want to put in place. Show result, data issues, the data quality, all that and sort of one seamless platform. So, at least for somebody like you or a person who’s overseeing many such experiments, they can decide, you know, which experiment is ready to take it forward to production and which one requires more pressure testing.
Shashank Garg 20:44
So Fady, what we realized also as the data scientists are sometimes not very good at being able to explain what the model does, and even when they do, because they’ve built the model, they’re so close, there is inherent bias in the way they’re gonna explain the model to a business user. So we sort of invested, create a framework around explainability and compliance. So the platform comes in built with those two things. And you can talk to the model, you can create different scenarios with the model, independent of the data scientist who’s actually trying to make the case, whether we should use this model. So just put an extra check on what you were saying that not every AI model is ready for prime time, especially coming in from the people who built it, and you have to have that oversight extremely, extremely important. Just around the topic of experimentation, Fady, is there anything else in your organization that you’re experimenting with that you’d like to share around this whole topic of data AI tech?
Fady Boctor 21:42
It’s interesting, I think we are in a place today, in addition to what we’ve talked about, there’s a significant amount of interoperability and data exchange. So, we think of TEFCA and QHIN and what’s happening with electronic health records, what’s happening with interoperable health information exchanges, the story continues on, there’s so much potential for experimentation. The beauty behind this experimentation is you’re building integration, models of integration. You’re not forming your own equation. You’re not forming your own, sort of, unproven model, you’re leveraging interoperable data availability to help your consumers, the patient, make decisions for themselves. So, we are excited to be a part of that experimentation process, ’cause like never before, the consumer’s data is so handy and available to them, well, why not leverage it and let it speak to their needs and let it customize solutions to wherever they wanna go, whatever they wanna treat. So we’re excited to be a part of that, sort of, development and that momentum, but that is definitely where we hope to experiment in the near future.
Shashank Garg 22:49
Fady, one more topic that I like to talk to executives about is this whole topic around change management and whenever you are disrupting, whether it’s internal or external change management and maybe we talk internal. So within your organization or your past experiences, whenever introducing, being more data driven or models, things that humans are not very used to, what has worked for you or not worked for you around the topic of change management adoption?
Fady Boctor 23:20
Well, the thing that jumps out at me most to that question is in our industry, because we’re so heavily governed by FDA and so many other subcommittees to sort of just the healthcare system, it’s important that we do not experiment on vital consumer engagement or vital healthcare and delivery. So with that said, one of the key areas is leverage existing technologies that have been refined, sharpened, and work phenomenally well, but maybe outside of your industry, maybe outside of your sphere, and bring them into your sphere. I think that’s where the exciting frontier comes in. And that also helps with change management because when you introduce something that’s new to your sphere, new to your landscape, but it’s been proven in its birthplace, you stand to have a much better, a platform of credibility, ’cause then now you’re interpreting how that proven technology can now be applied to your current sphere and how it could help drive momentum and progress in your current service area. So I think in terms of change management, go there first. Always look to see what phenomenal technologies already exist in other industries and see how they could apply and be customized and be transformed to your particular area. I think you’ll find tremendous mileage, tremendous amounts of options available to you with little disruptive, people are afraid of bringing on new things in new ways, change the new thing and just bring it into a new way.
Shashank Garg 24:50
Yeah. I like the thought process there. And you’re right, in a heavily regulated industry, people are gonna be that much more shy, so the easier or the smaller the change, at least the perception of the change, the smaller it is, the higher the chances of adoption. Makes sense. Fady, just a couple of more questions that I wanna get your thoughts on, one, if you look at broad pharma healthcare industry, what sort of technology trends that you’re most excited about going forward, anything that you’d like to share there?
Fady Boctor 25:28
Yeah. I’ll say that what I’m about to say speaks to an evolving healthcare marketplace, meaning we have about 120,000 physician shortage at hand, primary care specifically. And so what that said is a lot of the primary care and common reasons for a visit are starting to become difficult and more and more difficult to seek treatment for you, you’re on waiting lists, often you’re going to the CVS MinuteClinics, which don’t necessarily have a deep relationship with you, but they’re good in case of urgency, emergent care. But primary care is becoming more and more difficult. So what I see as an emerging trend that I’m excited about, are two things, one, first, the availability of our health records. So today I’m one of millions that has access to MyChart. MyChart gives me access to all of my healthcare records from, even from disparate institutions, I can see everything where I’ve been, what I’ve done, what’s been done, test results, appointments, everything right there on MyChart.
Fady Boctor 26:25
The consolidation of disparate health information has, is becoming more and more consolidated and more and more available. So that’s one key exciting factor. How does that feed into the next change? Empowering the consumer to be able to access that, to make critical decisions for their primary care. And with that, we’re a part of that expanded access by bringing more prescription medications to over the counter. So how do we integrate to that healthcare data, better, and equip the consumer to leveraging their healthcare data to make those critical decisions without a learned intermediary. And the story goes on. TEFCA and QHIN being developed over the last few years, and just the interoperability of health records in general, EHRs, Health Information Exchanges, they are maturing and they’re maturing quicker than we know maybe, than many of us realize.
Fady Boctor 27:14
That is an exciting new feature available in the healthcare marketplace, because now we’re starting to bring the consumer into the picture of healthcare delivery, where we’re starting to rely less and less potentially on primary care practitioners for the more transient, easy to diagnose, relatively easy to diagnose, relatively easy to treat conditions. And now we’re starting to empower consumers to make those decisions and seek those treatments for themself, leveraging their own custom healthcare data. I think we’re, it’s maturing quickly, but we’re still infants in this model. The future’s coming quick. I would say everybody should be on the lookout by the close of ’24, certainly by 2025, we should see some fairly significant leaps forward on those fronts. And it’s, that’s an exciting place to be.
Shashank Garg 28:03
And, well first of all, thank you for sharing these upcoming trends. And you’re right. And I was just thinking in, what’s going on in my mind is, so I know I track the data analytics and AI tech trends very closely, and what I’m starting to realize is the tech coming up so fast and combined with what you just said, the next 18 months, 24 months, can be very different for everybody in the life sciences healthcare industry. What sparked off as a, sort of, revolution or disruption with the whole generative AI launched less than 18 months ago, about 18 months ago, and we all gasped at the costs of training some of these large language models, and every quarter I’ve been tracking that very closely, it just keeps coming down.
Shashank Garg 29:08
It keeps coming down. And what we thought would be the cost to feed in all of your organizational data to a large language model, and to be able to talk to it in a meaningful narrative, in a narration, has come down a thousand times, has come a thousand times. So while the industry is progressing as you are talking, I think the tech is progressing very fast as well. And as soon as the cost starts coming down and we can get the right balance of experimentation and oversight, if I may call it, I think we are up for really, really cool things and transformation things in the coming future. Fady it has been fascinating talking to you. Any sort of last advice that you would want to leave our audience with?
Fady Boctor 29:57
Yeah. I’d love to, and I appreciate the time with Shashank, thank you, great topics of discussion. I would urge anyone who’s listened to our story to monitor PTPI, which is our taker on Nasdaq Petros Pharmaceuticals, and visit petrospharma.com, learn about our process, learn about our technology, the emerging technology and our current process in bringing the first prescription ED medication to potentially over the counter status by way of our technology, which could open the door for future portfolio assets. So, love the discussion. I think it’s the field that you and I are both in love with and passionate about and immersed in. And they should, people should tune in to see where we take it next.
Shashank Garg 30:34
Awesome. Fady, it’s been a pleasure having you on the show today. Your insights on AI, data change management, self-service, patient education and sort of the future of pharma and healthcare, incredibly valuable. Thank you so much for taking the time to share your experiences and expertise. And for our listeners, we hope you enjoyed this episode of the Intelligent Leader with me, Shashank Garg. If you liked what you heard, please consider sharing it with your network using the hashtag Intelligent Leader. And don’t forget to hit subscribe on our podcast by hitting the subscribe button. You wouldn’t wanna miss the next episode. Thanks for tuning in and see you next time.
Insights & Resources
The latest conversations from The Intelligent Leader.
Retail & CPGRetailers generate more data than ever, yet many struggle to turn it into meaningful business outcomes. This podcast explores the hidden challenges…
Retail & CPG 41:16 Ali Miller, VP of Ads at Instacart, and Shashank explore how data, AI, and collaboration are transforming retail media. Ali shares Instacart’s…
Life Sciences 29:56 In this episode, Kabir Patel and Shashank Garg explore the emergence of Agentic AI—intelligent systems capable of autonomous decision-making—and its transformative impact…
About this episode
Fady Boctor mentions streamlining clinical and research studies, physician profiling, anticipating consumer behavior, sales rep training with Generative AI personas, targeting prescribers and payers, and algorithms that help consumers decide whether a medication is appropriate for them.
Boctor says to educate the patient as the questions are asked, explaining why each question matters and offering tabs for more information, as a good doctor would. He says the FDA holds the company accountable for the public health value.
Boctor says AI cannot identify or correct its own bias, and in his example a vendor’s AI missed the implication of certain words in research study results even though it was mostly accurate. The FDA took notice, so there must be ongoing human governance and oversight.
Boctor recommends adapting technologies that have been refined and proven outside your industry rather than experimenting on vital healthcare engagement. Shashank Garg adds that the smaller the perceived change, the higher the chance of adoption.
Boctor says consolidated health records and interoperable information exchanges, including TEFCA and QHIN, let consumers use their own data to make decisions about primary care without a learned intermediary. He sees this maturing quickly.
Yes. The full speaker-labelled transcript of the 31-minute conversation is on this page, free to read with no form, with chapters that match the key moments.
The Enterprise AI & Data Transformation Partner. Delivering measurable business impact globally.
#1 on Gartner Peer Insights · 97.2% client retention
