AI for Healthcare Marketing – What You Need to Know, ft. Sujal Raju

Alan Tam (00:00):
Do you believe that AI should be used in healthcare? Is there a fear that AI will replace a lot of the work that physicians or healthcare marketers or business folks that are doing today?

Sujal Raju (00:13):
In my opinion, it’s inevitable. So let’s embrace it. There are things that we got to make sure that we learn from it and utilize, and I feel certainly there are things to be feared upon because we don’t know much yet as to what shape it’s going to take. I mean, this conference is about marketers, so we could use that and leverage that to provide better service for our consumers. And that’s all about making a brand that supports that consumer to get what they need at the right time, and then physicians and nurses with clinical researchers that could be automated. So there are several areas that I think healthcare can benefit.

Alan Tam (01:01):
Hello Healthcare. It’s inevitable nowadays not to be engaged with AI or artificial intelligence. From your newsfeeds to your last Amazon purchase, to your latest episode of your favorite show on Netflix, we’ve all interacted with knowingly or unknowingly AI. And AI has been used for at least a decade in healthcare on the clinical side, from physician dictation to image processing, to assisting with patient diagnosis. What about the nonclinical side of healthcare? Can and does AI have a role in helping healthcare marketers, strategists, and operators with their business initiatives and tasks at hand? Joining me here today is Sujal Raju, CEO, and founder of Enqbator. Enqbator is an award-winning digital solutions’ agency that has helped many leading health systems deliver world-class web content and experiences. Sujal, thank you so much for taking the time today and welcome to the podcast.

Sujal Raju (02:05):
It’s an honor, Alan, thank you so much for having me.

Alan Tam (02:07):
So sujal, you work with many different health systems. I’d love to learn more about what you guys do and how you’re helping these health systems.

Sujal Raju (02:16):
Absolutely. Well, first of all, thank you for the opportunity and I’m sure this session will be great. It’s the hardest topic around and I think there is enough to be learned still and enough that is already out there and people are adapting it at such an alarming rate that it is certainly interesting to see how it’s going to hit us. For us, Enqbator, we have been working with especially healthcare clients for over two decades now and working with them on their web initiatives initially. Then it went to the mobile and then mobile apps, and now mostly all digital is mostly mobile because everyone has gadgets and things that are talking to each other, especially with the autonomous cars. All the other things at your house or even at offices or in this case, the health systems, right? Everything is interacting with you, coming from wayfinding and things like that.

(03:08):
We have been part of this journey for, I would say about 10 years, we’ve been working with such artificial intelligence sort of applications from reading surveys and making analysis from it so that you can gather. If you think about social media and you have hundreds and hundreds of posts and comments, and then how do you have a single person, if you’ve employed someone as a social media expert to read all of that, make sense out of it and which ones to comment on or get a sense of the emotion of the person who is commenting and how do you respond to that both negative and positive? There is so much that could use AI and benefit from. So that’s primarily what our interest is and we research on that quite a bit.

Alan Tam (03:55):
That’s awesome. Help the audience understand. Help me understand what is AI, right? I think that people are not sure, what is it and how does it work?

Sujal Raju (04:07):
It’s something that… The AI acronym has been thrown around and everything that seems like dynamic, people say, “Oh, that’s AI.” Partly true, but I would say AI in lay terms is simply a machine that has learned enough and is able to make decisions like humans would. So it has lot of data that it analyzes and lots and lots of data that it can, and the more data you feed it, the more better it gets. And then it’s able to use the neural networks to be able to analyze the data from different layers at the same time and then be able to use that and give out predictions. And I would say decision making, like humans would. So that’s what is AI.

Alan Tam (04:54):
So do you believe that AI should be used in healthcare? Is there a fear that AI will replace a lot of the work that physicians or healthcare marketers or business folks that are doing today?

Sujal Raju (05:08):
In my opinion, it’s inevitable. It is going to come here, not just in healthcare, but everywhere, right? So let’s embrace it. There are things that we got to make sure that we learn from it and utilize, and I feel certainly there are things to be feared upon because we don’t know much yet as to what shape it’s going to take. Physicians, a lot of things that could benefit, right? This conference is about marketers. So we could use that and leverage that to provide better service for our consumers. And that’s all about making a brand that supports that consumer to get what they need at the right time. And then physicians and nurses with clinical researchers that could be automated. So there are several areas that I think healthcare can benefit.

Alan Tam (05:56):
Absolutely. I agree with you a hundred percent, but as we look at the application of AI, what are some of the boundaries and limitations within healthcare that you and your clients have come up against?

Sujal Raju (06:11):
So with the ChatGPT, which has come around and with the popularity, everyone’s just getting to it and be able to use it and not realizing what they should or should not do, right? So I think one of the first boundaries or limitations that I advise my friends in healthcare and leadership positions is that please let them be aware what it is, how to use it, and also make sure that they do not put any personal information out there just because it’s too tempting, especially the ones like imagery and things like that where it’s able to feed off an image that you provide and then say, “All right, enhance it or something like that.” It’s out there for anyone to take from there on. And some of these AI programs actually state that right there in their terms of you saying that, “Hey, we could use this to better our algorithms or use this for our own research,” which means that the data could be out there exposed.

(07:06):
So limitations are there, make sure that you use that for studies, if possible, employ that artificial intelligence systems within your system so that the boundaries are set so that data is not going outside, but you’re able to feed that from the data within and then you’re also able to get the research out of it or analysis out of it within the system so it’s not exposed to the outside. Having personal data exposed is the biggest risk, obviously with the privacy policies. I mean, one good thing in healthcare is that we have been exposed to HIPAA regulations for the longest time and we know, we understand what’s personal information, what are limitations and how we should or should not use that data. But then once you expose it to artificial intelligence, it just goes and unless you set it to say, “Hey, these are the things that you should not be saying or doing,” these are some of the things that we have to train it.

Alan Tam (08:03):
Absolutely. So is it safe, right? So I think the top of mind healthcare, especially on the nonclinical side, which I find fascinating, is I believe more risk-adverse than the clinical side.

Sujal Raju (08:16):
Correct.

Alan Tam (08:16):
So how do you overcome that? How do you help your friends and leaders in healthcare take that step and assume that risk, like this is okay to move forward with?

Sujal Raju (08:29):
So what I advise is use it for the softer things right now, for example, hey, you have comments from social media, right? All right, so let it analyze that and get you a summary, all right? Of, okay, what’s the general emotion of where this is heading? Is it positive or negative? And then react to it, right? Content writing is one place where you can say, all right, “I want it to give me ideas on topics or maybe give me an idea for the heading, the proper heading for doing this new story or things like that.”

(09:00):
Get some ideas like that, but don’t overuse it in areas that you don’t understand how it could produce content or use that content to be able to say things. It’s like one of the risks I say is are no different than think a person that you’ve employed, and if they go rogue, right? Basically it’s computer systems that have been taught to learn from whatever data you feed and then the algorithms or patterns that you’re saying this is how you want to produce the analysis. So if that has not been done right, it’s possible that it’s going to give out results that you possibly don’t want. So those are some of the risks.

Alan Tam (09:41):
That makes a lot of sense. So you’ve seen a lot of applications, you work with a lot of different health systems. Share with us what are the best examples of AI applications in healthcare today?

Sujal Raju (09:54):
I myself am very excited about all the different applications that AI could be used on. On the marketing side, I think hyper-personalization is a concept that I have been working on for a while. Think of Amazon, right? Amazon’s initial website had homepage and then you had navigation and other things, but now Amazon’s website is completely dynamic. The homepage that you get is completely different than the homepage that I get. It’s based off my patterns of what I have shopped on. And the philosophy before was like, Hey, make them stay on your site for longer, but Amazon does not care. Amazon only cares about how many products you can sell, right? So the less time you spend is fine with them, how many products are you buying? And they’re able to produce that completely to my taste and interest and it’s able to show me those and I’m able to make my purchases faster, which is completely fine with me.

(10:51):
I’m satisfied with that. It saves me time. On the clinical side, I feel like one of the research studies at MIT is what I take examples from, is that it’s lifesaving to find a cell that could be cancerous and detecting it as early as possible. And so AI has been used by MIT to study scans for the longest time all the records that they had or how much they could produce for this data study and find sort of like a pattern or data results which would show this cell should be looked at. And that’s life-saving because what could be analyzed as a cancerous cell months from now could move that patient into more of a danger zone and they require chemotherapy and other things.

(11:44):
Whereas if you were able to detect that so months ahead or years ahead, then you could take care of it in early stages and possibly cause a lot of healthcare, I would say benefit from these patients having healthier lifestyle and things like that that could avoid the chemotherapies and other advanced treatments. So I think that is one place, but imagine that kind of research and study using AI in all the different areas and I think it’ll improve healthcare.

Alan Tam (12:16):
Yeah, I really like that example. But I also do want to go back to what you mentioned, which is personalization. Share with me what are some specific examples that you’ve seen from a personalization perspective directly in healthcare. Like Amazon personalized homepages for each healthcare consumer. What are some examples in healthcare that utilize AI for that level of personalization?

Sujal Raju (12:40):
Yeah, I don’t have many examples of sites doing that, but we have done it and I’ll walk you through the trend of how it all started, right? So early 2000s, late ’90s, you see sites that are brochure-ware, right? It’s only talking about, hey, we have these departments at our hospital. It talks about that and has a picture. And that was it. Then came sites where you would have what they used to call MyHealth. So MyHealth was a way to personalize content so that when I log into the portal, I’m now able to see my results, my doctor, my location that I practice at, my forms that I need to fill and things like that. And then fast-forward now in late 2020s… Or sorry, early 2020s, late 2010s, and you see AI being used for taxonomy, which is the next trend where based on what you’re looking at, it then suggests you like, “Hey, these are the topics that you might be interested in.” That used to be in the right or the left rail.

(13:38):
And it would say, “Hey related topics.” And then it would get to it sort of like Amazon at that point saying like, “Hey, if you’re interested in this product, you might also be interested in this product.” But what we are working on right now is what we call as hyper-personalization is treating content in segments rather than a full topic. So for example, if you’re trying to create a homepage for let’s say cardiology or something oncology and major clinical services, the trend is all right, you have a homepage, possibly a few feature segments, and then you would take them to, okay, these are all the treatments and services and everything is structured, but still it is like one serving for all and everyone has to go through the same pattern and it may or may not suit someone who is looking for a very specific information. So the philosophy switches to say, you know what?

(14:29):
I’ll feed the content management system that’s capable of using AI, but segments of data. So not necessarily saying that, “Hey, this is for homepage,” but really, so that could be a welcome message, which could be one segment. And then anything that talks about my department, my physicians, these are all individual segments that as long as I’ve properly tagged it, titled it and things like that, AI will then put together all of that for me as Sujal or for you Alan, and produce information for you that is very relevant to what you’re looking for, right? It’ll be based off patents.

(15:02):
For example, like Google, right? When you make a Google search or when I make a Google search completely different, what results I’ll get is completely different than what you will get. It’s all based off the AI and it’s able to produce the page zero results, which is really giving you the answers, the [inaudible 00:15:18] snippets, giving you the answers right there. So I think that’s the progression for websites, mobile apps and mobile websites do the same. Provide segmented data that are stitched together and provided to the consumer based on their interests and their need at that time. There are a lot of parameters, not just interest, but also at that moment, what do I need?

Alan Tam (15:40):
Absolutely. And I do find that fascinating. I think that’s a fantastic example of that and would love to see more of that technology be implemented for healthcare. What’s fascinating is I still think that is very primitive compared to the other example that you shared on the clinical side where it’s like I’m using AI to detect whether this particular cell in this patient is cancerous. Why do you think there’s so much risk adversity on the business side of healthcare compared to the clinical sides? It kind of seems counterintuitive to me.

Sujal Raju (16:17):
Right. Right. It’s also because in healthcare there is a lot of gray areas of where someone thinks that, hey, this is okay to do. Whereas it turns out that it was not okay. Take an example of the Facebook Meta pixel. The whole of last year was all about damage control because no one realized that they were collecting this data and then someone sued a hospital system and then all of a sudden the legal teams everywhere are like, “Hey, stop everything. Just take off everything off from your website.” So I think taking that into AI where it’s very hard to understand it, right? Not many people understand it, even the biggest scientists don’t know where it’s heading, right? Because it’s able to make decision making like humans. So it all depends on what kind of data we’re going to feed it. And again, it’s open to anyone’s imagination of where we take this, right? And we see all these futuristic movies from 1980s, ’90s and even 2000s and you see flying cars and this and that.

(17:22):
Well, all of that, who knows what’s going to be possible? I mean, we imagined there was that back to the future 2020 or that they were talking about that era and we don’t have flying cars yet, right? But it’s accelerating at a much faster rate than we anticipated as soon as OpenAI became accessible through ChatGPT, because this is the first time or one of the first times where you now had a user interface to interact with AI. AI has been around and been advanced for the longest time, but this time it was made accessible through a user interface which people could now access and people are using it, the plugins and other things to put in a lot of applications and it’s going to be fascinating to see what those will come out.

(18:08):
So that being said, how does the legal team, the business team, cope up with it? To say that, “Hey, stop that or don’t do this, or let’s just do this.” That’s the hardest part that’s going to happen. New regulations have to come in and say like, “Hey, do not use for this, or if you do this…”

(18:26):
Just like how the car industry was before until when a lot of accidents started happening is when the regulations became stringent and now they’re better for us, even aircraft flying and all of that. So I feel like over here and every other application, every other industry will probably have regulations that will be created to put a safeguard rail around it.

Alan Tam (18:49):
Right. So that makes a lot of sense. Where do you see AI adoption going in healthcare and in the next 18 to 24 months?

Sujal Raju (19:00):
Different areas, I think. So for example, with 5G, which is allowing a lot more bandwidth, a lot more responsiveness and a lot more data that could be transferred at faster rates. I feel like a lot of internet of things, devices are going to come out, which means that as I walk into the hospital, I’m greeted, I don’t have to go to the front desk, but it already knows from my app that I’ve already given permission to interact with the hospital system to say like, “Hey, I have this appointment at the fourth floor, how do I get there?” Something is telling me through my earbuds or things like that, and I’m just able to do things faster. That’s at the clinic physically. But then you have research studies where using what’s called as merged reality, which is using virtual reality as well as augmented reality to do healthcare studies, right?

(19:53):
So you could have a specialist somewhere in, let’s say in Asia or in Europe, and they’re interacting with students here in North America or vice versa, and they’re able to sit like we are and in front of us, you have a 3D hologram of the heart and you’re studying it and you’re able to make decisions and research. Similarly, robotic surgeries could be happening remotely. A specialist from New York could do a surgery back in Australia, let’s say, and they’re able to do that, whereas other existing physicians or robots are helping with that. So it could be lifesaving, right? In marketing side, hyper-personalization is one of the things. The other would be providing data or e-commerce applications or signups that could happen online, not just on social media, but on your own website. How does that work out? Don’t know yet, but we all know that consumers are demanding it, right?

(20:52):
Look at how popular the electric cars have become. Look at how popular our home appliances have become that are AI based, right? So if you have… Running out of milk, the refrigerator is already talking to your shopping list and telling like, “okay, these are the things that are running out.” And then when you go to the grocery store or something, either they already have put everything together and ready for you to pick up or you enter the grocery store and it telling you like, “Okay, all the items in this store is at these locations,” so it’s giving you the best path to pick them up.

(21:21):
And then you check out, right? So similar experiences could be found in healthcare where you come in, come out, your pre-op and post-op treatments could look different and individualized, personalized for you, right? Because what physical therapy session could look like for someone may not be the best for someone else because they could have some other problems. Like they could have had a total knee replacement surgery done a couple of years ago. And now the hip replacement, so how does that look? It looks different than someone who it’s their first surgery. So lots and lots of areas that’s going to change.

Alan Tam (21:55):
Yeah, absolutely. And I think for healthcare, they don’t have to look very far because AI is actually already part of all our lives as healthcare consumers already. I know with my health system, day before my appointment, I’ll get appointment reminder. Once I walk onto campus, there’s the mobile check-in, which is really cool. The airline industry has had that for a very long time.

Sujal Raju (22:21):
Correct.

Alan Tam (22:22):
Right? And healthcare is just adopting it now. So I think there’s a lot of potential, a lot of examples, a lot of inspiration outside of healthcare that healthcare can turn to leverage that technology. Sujal, it’s been a amazing conversation. I really enjoyed this overview and I think you’ve painted a really good perspective in terms of how AI can be applied-

Sujal Raju (22:47):
My pleasure.

Alan Tam (22:47):
In healthcare, especially on the nonclinical side. Thank you so much for that. I’m sure many in the audience probably have a million more questions on technology and AI and how they can start adopting that. What’s the best way for folks to get ahold of you to continue the conversation?

Sujal Raju (23:07):
So you can find us on LinkedIn. My profile, its LinkedIn slash Sujal Raju is one of that, our website, Enqbator.com. We would love to hear from you. We are actually already working on a couple of research projects with some major health systems, and we would love to hear what’s your unique need and work with partners that I have in the industry that I worked with a lot of years. So we are certainly interested and knowing what your needs are, and I would love to work with you.

Alan Tam (23:35):
Wonderful. Sujal, thank you so much again. For those of you in the audience, this is just the tip of the iceberg when we’re talking about AI. And one thing that I agree 200% with what Sujal said is, we must embrace it. It’s coming, it’s here. Let’s understand it better and how it can help us in our everyday lives. So do reach out to Sujal on LinkedIn or Enqbator.com and until next time, hello.

Sujal Raju (24:04):
Thank you so much, Alan. Appreciate it.

Outro (24:06):
If you want more of the latest from healthcare’s thought leaders, subscribe using the button below or you can visit hellohealthcare.com to get updates directly in your email.