Breaking Healthcare’s Data Bottleneck w/ Shubh Sinha

Healthcare sits on mountains of valuable data, but compliance bottlenecks prevent organizations from unlocking its potential.

In this Executive Feature episode of Caretalk, Shubh Sinha, CEO and co-founder of Integral, explains how his company is changing that by treating data infrastructure and compliance as one unified system rather than separate processes.

🎙️⚕️ABOUT SHUBH SINHA
Shubh Sinha is CEO and Co-founder of Integral Privacy Technologies, where he's reimagining enterprise data infrastructure by unifying what was traditionally fragmented - the data stack and compliance stack. Under his leadership, Integral has pioneered a platform that certifies, transforms, and activates privacy-regulated data in hours—not months. This transparency-first philosophy has earned trust from top 5 pharma companies and Fortune 500 leaders who now process sensitive data at speeds that weren't possible before. Previously, Shubh led product management for LiveRamp's regulated data analysis, served as an angel investor in AI and health SaaS, and is a Forbes Technology Council member. At Integral, he champions the belief that intelligent software can conjoin what humans traditionally couldn't - making privacy-compliant innovation actually achievable at enterprise scale.

Integral Privacy Technologies is reimagining enterprise data infrastructure. We've unified what was traditionally fragmented - the data stack and compliance stack - into a single, intelligent platform. While others build compliance tools, we're building compliant-first data infrastructure that processes sensitive healthcare and consumer data at speeds that weren't possible before. Our AI-powered privacy engine doesn't just check boxes; it understands your specific use cases, data types, and regulatory environment to make smarter compliance decisions automatically. Fortune 500 leaders trust us because we deliver wild performance gains: 10x faster media log processing, dramatically reduced time-to-insight for HEOR use cases, and seamless unstructured data harmonization.

See your data's privacy transformation before committing a dollar on data with our Remediation Preview at www.useintegral.com

🎙️⚕️ABOUT CARETALK
CareTalk is a weekly podcast that provides an incisive, no B.S. view of the US healthcare industry. Join co-hosts John Driscoll (President U.S. Healthcare and EVP, Walgreens Boots Alliance) and David Williams (President, Health Business Group) as they debate the latest in US healthcare news, business and policy.

GET IN TOUCH
Follow CareTalk on LinkedIn
Become a CareTalk sponsor
Guest appearance requests
Visit us on the web
Subscribe to the CareTalk Newsletter

Transcript

David: 0:02

Healthcare is chockfull of valuable data, but so much of it sits unused, locked up in compliance bottlenecks, and that's particularly frustrating for innovators who confront the inevitable compromises between innovation and regulation. Well, today's guest says it just doesn't have to be that way. It is possible to unlock data and stay compliant. I can't wait to hear how. Welcome to Care Talk Executive features a series where we spotlight innovative companies and leaders working to advance in the healthcare field. I'm David Williams, president of Health Business Group. My guest today is Shubh Sinha, CEO, and co-founder of Integral, a company that's redefining how regulated industries like healthcare unlock the value of sensitive data. He joins us today to talk about how speed and compliance can finally go hand in hand. Shubh, welcome to Cure Talk.

Shubh: 0:59

Thank you. I'm excited to be here.

David: 1:02

You know, there is so much data out there in healthcare, and at the same time, I don't see all that much value being generated from it. I feel like I've been saying that same thing for a long time. Is that still true in 2025?

Shubh: 1:16

I think what you have been saying has historically been true, uh, but due to whether it's. AI or whether it's advancement in technology, specifically analytics technology in general. We are noticing the value exchange has pretty much gone up exponentially now, especially when you think about AI's ability to work with unstructured uh, data, which 80% of healthcare data is unstructured. And if you think about the, the daily lives of people like you and me in hospitals and care provider settings, you have tons and tons of that data being generated now. So I would say. That what you said is probably true up until 20 23, 20 24. But in 2025 it's, it's completely flipped. Where now, at least from our perch in the ecosystem, we're noticing a lot more adoption to take advantage of the, of the breadth and death in healthcare data. Not to say it's easy and not to say everybody's doing it today, but we are seeing a, a massive uptick now and we're excited to be helping and empowering that uptick as well by making that data usable so that it can be transacted upon.

David: 2:20

Well, I love to hear good news because, you know, it's not as common these days as, uh, one might like, so I'm looking forward to digging into that. Now, one of the parts to dig into, a little bit, into the nitty gritty here is that a lot of times people talk about data infrastructure on the one hand and compliance on the other, but if I understand right, you look at those as, uh, that is a kind of an outdated way to talk about it. What, why do you say that?

Shubh: 2:44

That's right, that's right. Before I started integral with my co-founder, I was previously at a, at a big tech company where I was helping large pharma and insurance enterprises leverage tons and tons of very sensitive medical data. And I saw firsthand that my products, which were data analysis, data ingestion, and insights products. They had to continuously accept either suboptimal data sets or had to wait 10 weeks prior to ingesting various data sets because the compliance timelines took so long or were so manual. And so I, at least firsthand saw it from my own incentives where I wanted to get that data as soon as possible and produce those insights to make those enterprises that are very, very. Uh, data hungry, uh, happy. But ultimately, and, and while we still made them somewhat happy, ultimately we always lost so many weeks, uh, and so much data fidelity to compliance that in my, in my role, uh, where I was working with the compliance consultants where I was. Owning my own product lines. And then where I was talking to the these enterprises, it was very clear that the separation of the data stack and the compliance stack led to a ton of friction. And so what we try to do is collapse those into one where you still equally respect both, but you treat them as if they are one, so that you can actually streamline the timelines and get that data to a much higher quality state. Because ultimately the end value here is speed and data quality to inform better product r and d, better distribution to people like you and me. Uh, and so by treating them together, you do notice a lot of inefficiencies. But most of the world, or mostly historical precedent, has been to treat them separately, which means you have conflicts, incentive, alignment, problems, uh, and so that's what we seek to solve here.

David: 4:30

You know, you mentioned like 10 weeks, which sounds like, you know, forever. Uh, as far as I'm concerned, especially in these, you know, sometimes you take like 10 milliseconds is too long. I might navigate away from a webpage. So I think that sometimes, you know, people wait 10 weeks if they have to, but I, but I also think you end up with just a lot of data that's just sitting on the shelf. You say like, I, I can't wait 10 weeks. I can't put that investment in. Do you see that today? Are there, are there examples of sort of valuable data that gets kinda wasted and just sitting around?

Shubh: 4:58

That's right. That's right. And, and in my previous job where I was responsible for utilizing that data, but ultimately I, I couldn't do anything until it was compliant, uh, to make sure we followed all the regulatory rules I saw firsthand. We would leave a ton of very valuable data on the table in, in favor of getting what could be done in nine weeks versus 12 weeks, or five weeks versus 10 weeks. And so, you know, the easier data sets, quote unquote, or the less, uh, less rich ones are easier to handle. And so those end up getting done more quickly, at least with the incumbent system. And so you end up using those'cause speed at some point, uh, you know, makes the difference. But the high quality data assets that we now streamline, uh, and that we now automate, uh, from a, from a, a compliance analysis and data perspective. It's very clear there's a night and day difference when you're able to use a very high quality rich data set on, on consumers, uh, versus being able to use whatever's easy to grab and off the shelf. And so I think historically the precedent was to go after some of that less rich information because it would get through compliance in a more streamlined way. Five weeks versus 10 weeks. But you're right, you did leave a lot of valuable data on the table that if you could make use of makes difference, makes a big difference on the p and l at an enterprise level, makes a big difference on the product tailoring and customization for consumers who ultimately are the recipients of these medical products that make a big difference in their lives. So the, the more tightly you can curate the data while. Balancing privacy and compliance, the better it is for everybody. But ultimately, yeah, you do wanna leverage the, the rich data sets. Just they're, they're off the shelf or they're on the shelf, I should say, for, for quite a bit of time.

David: 6:39

Got it. So I can see why, you know, in your, in your prior role, and for people that have to deal with these data, it, it could be frustrating to wait and maybe sometimes you don't get the exactly the right thing, but in terms of putting together a value proposition for what you're doing. How do you think about how delays in compliance processes end up translating into more measurable costs for, you know, those on the healthcare analytics and research side?

Shubh: 7:05

Definitely. And, and the way we think about our value or our, our ROI in the ecosystem is, is really threefold because when you work with some of the largest enterprises in the world, yes, you have a responsibility to that team. They're your customer. But ultimately these enterprises, you know, they're basing a ton of very, very high. High risk, high important, uh, decisions off of the data sets that you're deploying, or I guess in our case that we're deploying. And so we take it very seriously that not only do we need to make our customer happy, but these, these customers, uh, for us, these large enterprises, they're building products for consumers. Like you and me that we will consume. It has a real, uh, really a chain effect in a way where our job doesn't just stop when we deploy that data set. Ultimately the quality and what's acted upon there, uh, needs to be robust. Uh, and so we take it very seriously to the point where we measure ROI in a couple of different ways, which is speed for sure. Being able to do it in a few days as opposed to a few weeks or, or a few months. Provides immediate utility on its own because it increases the rate of experimentation and comfortability from these data scientists that are developing these groundbreaking products or are developing groundbreaking, uh, distribution strategies. The other part is data quality. Ultimately, a lot of the data sets that we are deploying for these large enterprises are for distribution purposes or r and d purposes. In which case, the more tightly it can be curated, the better these products are for people like you and me. And ultimately that is a very big value driver for people's everyday lives. People's trust with these enterprises and then. A little bit, uh, related to that, we definitely think about how are we balancing trust and privacy with the consumer, leveraging consumer data. You know, if you, if you just said that sentence on its own, uh, it, it kind of gets a shock. But if you do it from a compliance and privacy perspective where you can walk that line between data, utility, and trust, ultimately it's a better bridge for both. And so we take it very seriously to, to uphold that, uh, trust bridge, so to speak, from a consumer perspective as well, because. It should be win-win. And I think news articles or headlines can make it seem like a loss, uh, in many ways. And our goal is to make sure that is never true where Integral is deployed.

David: 9:20

Those of us in healthcare tend to think of it as being unique and there are a number of unique elements, you know, related to something very personal, someone's health. But there are other regulated industries, I'm thinking like finance or aerospace. Are there parallels between, uh, you know, some of the issues that you're talking about here in healthcare and what you see in those industries or others?

Shubh: 9:40

Definitely, definitely. And I've been very fortunate to meet people from other industries who. They say when you look at something completely different, you end up seeing similarities. So whether it's my friends and colleagues in the logistics space or the finance space, uh, what I'll say other regulated spaces or my previous employer, you know, we had a healthcare organization within the broader company, but the broader company itself was not a healthcare company, which meant that I had exposure to CPG retail media. Uh, there's a whole different. Note about how I believe healthcare data will affect all of those. Uh, but for the current purposes, you know, the non-regulated data industries, you see how fast they move with their data. You see how they're able to be so agile and ultimately churn out more products or have more experiments, more shocks on goal to develop better products or people like you and me. So when I, especially when I came into healthcare because I organically came in, uh, and discovered it for myself rather than, you know, kinda like a straight trajectory, I, I realized. The data practices or the data, uh, or, uh, getting value from that data. It's, it's no different than a non-regulated industry. A data scientist is a data scientist, uh, and a product is a product in a way. Obviously the compliance and regulatory hurdles that present themselves are a challenge and an opportunity for healthcare, whereas they're not in non-regulated industries. And so one thing I always think of is are we reinventing the wheel or are we adapting the wheel for our space? And ultimately, I believe that while we're working on very unique, very interesting stuff, we're also trying to make sure we can bring the speed that is present in non-regulated industries. Into regulated industries. And I also see the, the parallels in finance or some of these other companies that we work with outside of the core pharma and insurance verticals, uh, whether it's in tech, whether it's in media, you know, that they're utilizing healthcare data, which has its own regulatory hurdles and whatnot, but. I can just see that their, their data infrastructures are set up to be so much more agile in a way. So there's definitely some parallels for sure. But I think healthcare in and of itself, because it's been so heavily regulated on so many different sides, you see a large opportunity for creating a new type of data infrastructure that can introduce tons of efficiencies in different corners, that, that all add up to be pretty, pretty big.

David: 11:57

So going into, into healthcare and talking about hipaa, I believe you've talked about HIPAA safe harbor and said it should be a floor and not the ceiling. Can you unpack that a little bit? First of all, what is the HIPAA safe harbor? What does that mean and and why is this an important distinction between whether it's a floor or a ceiling?

Shubh: 12:15

Definitely. So at a high level, as I'm sure you know, and, and your audience knows, HIPAA is the, uh, regulation that governs healthcare data utilization. Uh, and so if you are an enterprise, whether you're 5,000 people or whether you're two people as a company, you have to abide by HIPAA when you're utilizing very sensitive, protected health information. So hipaa, safe Harbor and HIPAA expert determination are two compliance frameworks. Uh, we treat them as almost as configurations in our software. They're two compliance frameworks to make sure that a set of data, a set of input data. It can go through a transformation process. It can go through a compliance process and come out the other end, ready to be analyzed, ready to be used for model training, whatever the use case might be. But the, the whole point of that transformation compliance process is to ensure that you have sufficiently reduced the risk of identifying a person like you and me, because the regulation outlines that primarily for protection. So definitely there for a good reason. That being said, how they do it is quite different. HIPAA Safe Harbor has a very strict, uh, scrubbing. It's almost like a set of scrubbing rules, uh, in addition to many other, uh, very abso isms is maybe the right way to say it, versus HIPAA expert determination, which is, you know, we support both, but we really specialize in the expert determination. It's more contoured, it's more fit for purpose in the sense that it aims to reduce the risk all the same. So we're, you know, nobody is, uh, saying any risk is okay, but it aims to reduce risk in a more contoured way, such that if you are an enterprise, for example, looking to study. Vaccine distribution by geography. Maybe you can keep in more geographic elements and take away from other profiles of the dataset. So maybe you knock off demographic and financial variables in that dataset and you create like a very, very contour geographic dataset and expert determination allows you to do that contouring while reducing the risk, whereas HIPAA Safe Harbor is. One size fits all. So it would treat that data set as the same as maybe any other type of healthcare data set in a way so that it's a, it's almost like a scalpel versus a hatchet is what we like to say. Colloquially and the scalpel allows you to be a bit more surgical.

David: 14:22

Okay. That's good. Yeah, I can see this is a, a good concept to wrestle with. So if I just wanna make sure I understood your example. So you have a rich data set to begin with, and you wanna reduce very, to a very large degree, the possibility that someone's gonna be revealed. But if you say, what's your objective for the use of this data? And if it's to understand, let's say, maybe fine grain geographic differences, perhaps because it relates to, uh, what retail stores you're gonna offer it in or, or some, some element like that, you can focus on that by kind of blurring some of the other elements that might, might be more identifying, if you got it down to the specific level, but that aren't really as important for the analysis, is that, have I understood that

Shubh: 15:01

right? Exactly. It gives you that flexibility to blur such that you can focus on what you need to.

David: 15:08

Got it. So I would, I don't know what you can offer in terms of a customer example, but I'd be interested, I get this conceptually, but is there an example you could provide of a customer who went from maybe being stuck with the previous issues of, Hey, I've gotta be compliant, but I can't really achieve my objective to becoming successful using the integral platform.

Shubh: 15:28

Definitely, definitely. And, and maybe just to cite a, a high level example, you know, we work with a lot of the top pharma companies, the top insurance companies, media, tech companies, et cetera. And what we find is a recurring theme is that when they switch over that, that very first time from a consultant, uh, to start deploying their data sets via. There's the speed, ROI. So we had a pharma customer in particular that was only able to activate maybe two to three data sets a year because the reviews would take eight to nine weeks, if not 12 weeks per dataset. And so from, from a clear ROI perspective, you know, they had a ton of data. They had, you know. They say pharma is cash rich and data poor, uh, in many ways. And so they had ton of data that they had bought, uh, and that they wanted to be able to connect in order to measure the distribution strategies they were experimenting with. But because they can only activate a portion of that data. They were always left a little bit blind to some of the experiments and the results therein. So the most clear ROI when, when this customer switched over to us, we got them set up on our real-time pipelines that we ended up deploying their dataset that was historical. So everything they had done before with the consultants and then they ultimately ended up deploying a dataset approximately once every month, uh, as if I recall. And so from that, you have a clear ROI on just the data budget and how much of it you recouped. The speed in and of itself because the data scientists were effectively fed every month with the latest and greatest data. And then the data quality was probably the, the more, uh, uh, clandestine, but still very, very powerful value proposition there where we were able to. Introduce more variables than they previously could because we could take the time to be more contoured in a way. And so from a clear ROI perspective, you have not only recouping the data purchase budget because you're ultimately in the red until you activate it all, and then speed the data quality and then the overall org satisfaction. Uh, this, you know, from my own experience, when you're activating these types of data sets. The compliance side, the consulting version of it, the more manual version of it can introduce, earn risk or you know, risk with the actual products that you're using. And so by, by making that a process streamlined, it actually has second and third order ripple effects where our data partners, in addition to the data purchaser or analytics partners, in addition to our day of purchaser customer, they're all very happy when we tend to get involved.

David: 17:51

Got it. So. I don't know how you can describe this. I understand it qualitatively, but I don't know if there's a way to go beyond that, which is, I, I understand how these data sets can become more useful when you, you process 'em with your platform, how much more useful are they? Is it like twice as useful? I understand the speed part, but like the, ultimately the dataset they, they get,

Shubh: 18:14

yeah, so from a numbers perspective. We tend to, uh, it's because we deal with so many different data modalities. It's hard to have one average for what we introduce as a quality variable. But I would say on average we can introduce anywhere from 30% for some of the, more from the, some of the more, uh, rare data sets, 30%. Up to 70 or 80% more data retention. Meaning if you had a hundred rows in your data, if you were losing 80 of it before, we can bring that, you know, we can bring that up to 80, uh, from 20. And so the, that, those types of metrics we try to compute internally because we're. Trying to focus on making sure that that data ROI is as solid as possible. Uh, so from a usability perspective, maybe one way to think about it is not going or going through the compliance chopping block, but not actually getting a whole lot of your data set chopped up because you were able to be more contoured and tailored. Uh, I think the, the other way we think about the usability though is that the rate of iteration, once you, uh, once you enter into a compliance process, you know, they call it governance because it governs what you can do. So sometimes these data scientists write their job is not to play compliance analysts. That's our job. And so what they end up doing is putting a data set through our platform with the variable preferences they might want. And then because compliance either allows for most of those variables or it says, Hey, this particular thing might be blocked, uh, until you look at this, this, and this, it almost gives them new ways to think about the composition of the dataset. And so from a usability perspective. More on the qualitative side of course, but we've had data scientists tell us the rate of iteration is so high with integral that maybe the thing that we thought we wanted. When we saw it, we realized we actually wanted something else. But because integral makes that loop easy enough, uh, and seamless enough to do. They are happy to enter that loop, which goes back to one of my earlier points. A data scientist is a data scientist, so the goal is to introduce a lot of speed so that even if you think you know what you want, you can still confirm and on the off chance you might want something new. You can have that loop for yourself to make sure that you're confident.

David: 20:16

Got it. So I can see how the speed. It's useful to reduce the cycle time and you can have more iterations and then you have more of a chance to say, to learn, oh, maybe I wanted this and not that. Which you can't do if you have to wait, you know, so long in between it. So I think that's one point, and I, I can see the value, um, of that. Another thing that I've noticed, I I, I've seen it elsewhere, not necessarily in data science. I wanna ask you about it, is that sometimes in a big company, like in a pharma company. They may presume that, oh, you know, compliance is not gonna allow this, or regulatory is not gonna allow it, and they kind of self-censor and they don't ask for it as a result. It do. We have that same phenomenon here where people say, I'd really like to understand, you know, some element, but you know, compliance is never gonna allow it. So they don't ask for it, but they might actually now ask for it if they're using integral.

Shubh: 21:03

Definitely. And this was something that we or I had organically seen in my last role, but as we. Brought on more and more customers at Integral I, I saw it more and more often that if you can make the time to deploy these data sets very quick. The usability becomes quite high. These companies, which have tons of data, purchase budgets, uh, within them will go out and get more data, whether it's from, you know, data providers, new types of data sources, whatever that might be. And so the, the hunger and the appetite to ingest and analyze and activate very sensitive data is quite high. In which case, when we made the timelines to deploy much more, uh, much more conducive to experimentation, more data started getting purchased. Then our customers came to us and, and we detected some of this from our, from our own experiences. They said, how can you get involved from a feasibility perspective, not just from a deployment perspective? So how can you. Almost forecast what the impact of compliance will be on a dataset prior to purchase so that I can be more informed when I'm making a purchase and I can know what I, what I might have to lose, or what I might have to sacrifice or what I have to contour. And so we have a product offering called remediation preview or privacy preview, or sometimes called transformation preview, but it's almost a look around the corner with regards to what you're thinking about and you can. You can use it for your process. So you can type in, this is what I'm planning to do with these types of data sets. Or you can upload actual sample data to our platform, whatever the, the inputs there in and we will give you a high fidelity, never a guarantee, but it's a high fidelity confidence. Uh, uh, prediction as to what you can expect from a risk perspective, and that allows you to either purchase that data set, allows you to ask more informed questions as a data buyer. And, and our data provider partners love it because they're actually able to create what is an, like a more informed profile. Uh, and so everybody ends up winning when the, the flashlight gets flashed a bit earlier in the process. And so that's something that we see a lot. Uh, so, you know, I'll categorize it as. We are still focused on the deployment side, and that's still where a lot of our resources go. But we realize we present a ton of value and we become quite close with our customer if we can be on the feasibility side and at the natural extension, because we want to own the end to end for any data project that has compliance, uh, and regulatory hurdles. And so that's where we, we've started to dive in a bit more and noticing that if we can be a partner early on, it provides a ton of transparency and outsized value.

David: 23:36

That's great. I, I had a question which you've kind of preempted and, and answered, which is about, you know, you'd mentioned that before that organizations can see how compliance is gonna impact their data before they make a commitment. Then I wondered if it was feasible and now I'm understanding, you know, how that is. And then my follow on question to that was going to be about whether the data suppliers themselves make an adjustment based on the fact that integrals around and can offer that type of insight.

Shubh: 24:03

Definitely, and that's where incentive alignment is very important for our go to market. In particular, if you think about the incumbents, when a set of data becomes ratcheted down to where it's maybe. 30 or 40 or 50% of what it was. The data purchaser is not happy because they are not getting as much as they paid for. And the data provider is also not happy because they're selling a product to a customer that they want them to return and buy more. And ultimately they have no control over what's getting, or they have very little control over what's getting across the line. And so in this way, both, both sides get to make sure that their values and their incentives are quite aligned because there's a more contour process available and. It's almost as simple as the data provider wants to get more of their data through, and the data purchaser wants to activate more of that, and so this platform gives them both a way to transact on that.

David: 24:56

Got it. So I don't know if there's anything in here, but I was wondering, um, because I've seen other technologies that are out there and are useful people start to use them. And although clearly you have a lot of experience before putting the products out in the market, a lot of ideas. Sometimes a customer may use a product in a way that you didn't think about. Now I've already got the speed and also the feasibility elements of it. But are customers doing anything else with your technology maybe that you did not anticipate?

Shubh: 25:25

A direct technological application, but I did recently hear of a few instances when there were internal qbr, quarter business reviews going on, and our platform, either a screenshot of it or the actual platform itself was presented. And, and to be clear, we are not at these meetings. It's a purely internal with executives. But the platform is presented as part of the overall review process because these data budgets are so large that the data practices and the, the data organizations get reviewed and integral is seen as a key mission critical infrastructure for enabling that data spend to be recouped. And so, you know, not, not a loss, but actually a gain. And, and so from, from my perspective, you know, I, I always thought the platform is geared towards. Data scientists or those who are gearing, uh, gearing up for harnessing data and are usually the ones in the weeds. We definitely have executive stakeholders and whatnot in the platform. Observers, but you can, you can see they're, they're rarely in it, uh, unless there's a problem. Uh, so it's a good thing that they're rarely in it. And, and so when I, when I heard about this happening at not just one, two, or three, but like quite a few of our biggest customers, I was thinking, wow, it's so interesting to take what is a technological and data platform, uh, and of course a compliance platform, but almost use it as part of your presentation for. Enabling a data budget. Like obviously that's, that's the goal there, but it being used as a presentation layer, uh, was just quite surprising. Uh, so, uh, I would say that's, that's a surprising one. Even though it's not the tech itself, it's just the fact that it appears to, uh, it appears at such a high level executive style meeting. It means it must be introducing quality of life and it must be impacting some KPIs for very high up people.

David: 27:04

Nice. That's what we like to see, right? Yeah. Yeah. It's a good, it's a good surprise. Yeah. Shelby, you mentioned the word transparency before. How does transparency play into what you do in terms of, you know, building trust? A lot of these executives at some of these companies, a lot of large companies, you serve a very risk averse. So how, how does transparency fit in there?

Shubh: 27:26

Definitely. So earlier, I think it was one of my first points I mentioned, we try to collapse the data stack and the compliance stack into one, and I think that remains a part of our ethos. That being said internally is very important that we always remember that we have the positioning of a compliance tool and a data tool somewhat separately. In times where you'll go through a compliance vendor review, which is more aggressive than a data vendor review. So it's almost like this, uh, uh, like this dichotomy of yes, we are seamlessly combining them into one, and compliance is a byproduct of our data infrastructure. But also it's very important to build that trust because you, I always joke, you need to trust your compliance people. Uh, and so that's, you know, that's a very, very important part of it. You, you can build trust with your data people, uh, but you need to have almost default trust with your compliance people because the risks are so high and the, uh. The potential damage is so high. And, and so from that perspective, we always treat it as an ethos. You know, build that trust early on, such that when we're, when we're performing and we're doing well, it's from, it's from a place of, okay, I know, you know, I know these people and they're doing well. That's great. So trust is a, is a massive part of it. And the transparency on the other side is, is baked into the tech where I always say during calls when I'm talking to customers or potential customers, we, we almost write out our secret sauce, uh, in many parts of our platform. And, and obviously we're not giving it all away, but we do tend to be on the more transparent side with our audit trails and whatnot. And you can dissect how we do what we do because ultimately we realize that's a key, key part of trust there. And so to, to a degree. Transparency is probably one of the top three words used in the company to describe how we shape our products, how we shape our sales process, our customer implementation and onboarding. And so it's very much so baked in because ultimately we are not only enabling data sets, we're not only slicing and dicing data, we are doing it in a way that reduces regulatory risk. And so for that trust is probably the number one thing.

David: 29:30

You know, there's a lot going on in Washington DC these days. Um, with the change of administration, a lot of it having to do with regulation, and I think sometimes just, um, you know, described as like deregulation or dismantling agencies and so on, and there's some of that. There's also, I think, a realization that some regulations may be outdated, that were meant for, you know, certain purpose and like a lot of regulations related to privacy are from before the digital era. Nevermind. You know, ai, and I was struck by what you said before about how the data providers are influenced by what you're doing. I'm wondering whether there's an opportunity to influence how the regulators themselves view innovation, or whether that's beyond a scope of what you consider.

Shubh: 30:17

Yeah, definitely. And, and it is an ambition of mine that our company can at some point influence some of these regulations, but more from a talk to talk, talk the talk, and walk the walk perspective rather than just talking. Uh, because I like to think that our technology can show that if you have regulation. Uh, you know, it can be a good thing. Uh, you know, in general, I think deregulation and all of these things, you know, they get talked about as a way of introducing efficiency. But especially in healthcare, when you're dealing with such protected information and there are any number of incidents in the past 20 years that have seen information leaks and whatnot, and you realize just how these leaks can be so. So significant where, you know, not to reduce the significance of any other type of leak, but you know, when non-healthcare data providers get leaks, it's, it's still very sensitive information. But when you're in, you know, when you're in a healthcare data provider's office or whatnot, you're pretty much. Saying all the information that is unique to you, uh, so that you can get them the highest, uh, quality care. And so if that information leaks, it's, it's usually the worst type of, or the most severe type of information to leak. And so when you think about deregulation and all of that, and potentially introducing the hi higher chance of information getting out there, that I think is when it, it should go, uh, the opposite way or it should at least kind of stay where it's at, where there's a safeguard in place. And, and so my ambition would be to show that you can take regulation almost. Introduce technology as like a way to streamline it and automate its implementation such that you have. Equal, you can say these sentences out loud. Regulation is allowing for safety. Regulation is not slowing us down. I think that's maybe the, those two sentences are, are opposites today. Uh, what, and I think this is true of probably, you know, any type of administration in the office where it's, there's a disconnect between the technology on the ground and the policies, uh, being made. And so my, my goal would be one day to be able to join that, join that gap in a way. Uh, and so that, that I think represents a personal ambition of mine.

David: 32:19

That's great. I understand now from, uh, this discussion about how there's value and ROI for those folks that are using and supervising, uh, you know, the activities that Integral is selling into. I wonder though, at the top level, you know, you sort of hinted at there's some KPIs there. What's the bottom line impact for pharma companies or health systems overall, do you think? It's mostly about. Cost savings, is it about revenue growth? Is there something else? What's your, what's your sense of that?

Shubh: 32:53

Yeah, and, and maybe just to set the context, we're generally involved in the commercial side, meaning there's a product out in market. We have some involvement on the r and d side, but we, we see a bigger splash and, and a, you know, more, uh, let's say more acute pain point on the commercial side. So that's where we've put our foothold. And so when you think about the context of a product already being out in market, or a service being out in the market. Being able to boost its visibility and being able to understand why its visibility is being boosted or why people are buying it or not buying it is such a. Fundamental value to driving revenue, reducing costs. I would say maybe all the above of what you just said. That being a foundational layer in that, like that's where we truly focus. And ultimately, I think if you remove it one step away from the enterprise, if you think about like a potential prescription being out in market and introducing its awareness for a very, and maybe it's for like a rare disease, for example. You have a very real impact where the more data that goes in, if it's leveraged properly, it's the more distribution that goes out. Which of course, the enterprise seed is a win. But then ultimately people who may not have known about it, but have that rare disease or have that condition can actually get, get awareness and you know, it's up to them if they want to, uh, you know, take that or not. Uh, that particular prescription. The fact that they know about it in a very targeted way that that speaks to them and their profile, but in a way that's also very complimentary. That I think is the, is the perfect handshake, uh, when you're thinking about a commercial healthcare outcome. And I, I wouldn't say that goes to, you know, just prescriptions itself. Like there's awareness campaigns around te getting tested for certain types of things. There's, I would say largely there's a new. Health wave, uh, in the last five to 10 years that's been taking over the population, which is very good. And so those, even those types of awareness, uh, campaigns and whatnot, you know, people don't realize their, the reliance of that campaign on, or that success of that campaign is relied on by the underlying data. And so that's, that's where we see, uh, kind of like a big impact from our side.

David: 34:53

Great. So you've, you've answered my question in an even bigger way than I intended it. It was supposed to be a, an, you know, kind of an overarching question, but if I understand it, basically you're saying if we look at the whole ecosystem, the essence here is that that end. Patient, consumer, what have you, has a need, uh, which can be addressed if you've got the right data to get to that person with the right message. And if you do that, then presumably since it's in healthcare and we're talking about regulated products, probably gonna help them. They're gonna be healthier, et cetera, they'll be more productive and all those good things. The way to get that is. Companies or organizations selling product or service, you're helping those organizations to identify, uh, the right folks and get the message across well by handling the underlying data and sort of erasing the challenge between compliance, um, and innovation, um, and use. And that way you really, at the fundamental level of adding value to those organizations, which your, your customers. So that, that's pretty, that's a pretty good spot to be in. I'd say

Shubh: 35:51

so.

David: 35:52

Yeah. Good. I like it. I like it. Okay, so let's just say, as long as we're talking big, I would just ask you a little bit about the regulatory environment. We, I, I'll kind of go back a little bit to the question about, you know, what influence you might wanna have eventually, but if you had, um, if you could make some sort of a change in the regulatory environment that would allow more value to be unlocked from healthcare data. Are there any kind of specific things that you have in mind on your wishlist and you know, do you see any of these things happening?

Shubh: 36:22

It's a, it's a good question because I generally tend to think about it from the tech angle, uh, because you have tons of old health systems, tons of legacy types of systems that are implemented and they don't share data with each other, whether due to tech reasons or incentive reasons. And so I always think about, you know, if you had an end, this is to a degree the goal of integral. If you had this interoperable interoperability infrastructure that can plug into. Any data source and any data destination, uh, which is what we do today. And, but if we expanded that to go all the way down and up the data supply chain healthcare, you know, you could enable what is the, the one record for all in a way, or you could enable the, uh, very seamless understanding of what's going on in the health system. It's almost like very, very transparent look because it's all unified under one, one. Uh. Type of data infrastructure, but that's from the tech angle, of course. Uh, because the regulatory side, you know, we think about that as our opportunity to bake into software. I would say if I'm just kind of thinking about on the spot about the regulatory side, if I think about like things like HIPAA or whatnot, the, even the, the frameworks that we talked about, expert determination and hipaa, safe harbor and whatnot, those are. Of course very important frameworks, but they're also relatively general in the sense that they, they speak to all protected health information, uh, that is collected. So anything from you, uh, and your cholesterol rating to a voice, audio of you, uh, describing a reaction to a particular prescription you took, or something like that at a call center. And so it's a pretty wide set of data modalities that the current compliance rules speak to. So if I could think about maybe like a new type of regulation or introducing more, uh, but in a way that is enabling success, not disabling it by any means. It's almost like how do you think about the data modalities as opposed to just having a one size fits all? Because we, I mean, on our end, we do have a different process for treating audio data than we do a prescription receipt, for example. And, and that is necessary because they are very different types of data sets that contain very different values. And so we've almost taken it upon ourselves to make sure we have different levels of configurations. But that, that I think, could be an important regulation because on one end you do have the data purchasing appetite increasing quite a bit, which is why we, you know, taken our customers. But the other side, which is true with or without Integral, I think there's a ton of new proprietary data assets coming to market because new types of wellness apps produce this data as an exhaust, and so they end up monetizing it or they end up distributing it. And so if you think about all the new, it's almost like two sides. You have data purchasers and all the new data providers. There's going to be so many modalities that a one size fits all will eventually break.

David: 39:03

Got it. Well, I have one more question for you, and I'm asking you to take a look at your crystal ball if you may have one handy. And, um, if we look ahead, you know, pick your, pick your timeframe, but let's say three years, five years. Are we going to be in a different spot in terms of having, you know, much more data available, really used, uh, quick iterations and so on? Or, you know, you, me, you mentioned there've been some changes in the last couple years. Do you see that change continuing and improving, or are we just sort of at a new plateau?

Shubh: 39:35

Yeah, definitely. And something I spend some time thinking about, I would say we're gonna be in a totally different spot in the sense that. I can't remember the exact number, but it's such a large number that represents all healthcare data today. It's like zetabytes or so, something along those lines. It's a, it's one of those types of types of words, but with not only because, but especially with ai, you see these models, you see these very industry specific applications. The key to success there is going to be the quantity of data that these models can train on in order to be more effective. And so from my perspective, with AI coming in, the amount of data that these enterprises need to actually leverage that AI to be as effective as a team of very skilled humans, it's just so large. I think we haven't even scratched the surface yet, and that data that is so large or like that, that that large appetite's going to be. Fed by proprietary data, not what's on the open internet, especially in healthcare. And so you need a channel to enable proprietary data assets to go into AI models or to go into whatever the use case is, to be able to actually utilize the full value of it, specifically in healthcare, where the rate of error, it has to be small. And so I think we're gonna be in a totally different ball game because. Data providers will come to market. There's going to be entire data economies that start happening, uh, by modality. So like, you know, an entire voice data ecosystem, entire written notes, data ecosystem. I think all these things are going to happen, which is a big bet for Integral as well. We believe all these things are gonna happen, such that AI can actually be fully leveraged in healthcare, where the rate of error has to be so low.

David: 41:20

Great. Well, that's it for another episode of Care Talk Executive Features. My guest today has been Shubh Sinha, CEO, and co-founder of Integral. Integral is a company that's redefining how regulated industries, especially healthcare, unlock the value of sensitive data. I'm David Williams, president of Health Business Group. If you like what you heard, I hope that you're going to subscribe on your favorite service. Shubh, thanks so much for joining me today. Thank you for having me. I really appreciate the conversation.