Evolving the Enterprise

The Future of Healthcare: Data Integration and AI Insights with Ankush Khona, Data and AI advisor at AuxoAI

• SnapLogic • Season 3 • Episode 5

Share episode

Welcome to this episode of the Evolving the Enterprise Podcast, where we delve into the intersection of data, artificial intelligence, and healthcare. Our guest, Ankush Khona, a data and AI advisor at AuxoAI, shares his insights into the transformative power of data and AI in healthcare.

Key Takeaways:

  • Aligning Data Strategy with Business Objectives: Ankush emphasizes the importance of a data strategy that aligns seamlessly with the overarching business strategy, ensuring that healthcare organizations can meet their desired outcomes effectively.
  • Embracing Modular Infrastructure: He advocates for a modular and flexible infrastructure that can adapt to the evolving needs of the healthcare industry, enabling organizations to integrate new technologies and data sources seamlessly.
  • Navigating Data Privacy and Security: Ankush highlights the critical importance of data privacy and security in healthcare, outlining the need for robust frameworks and standards to protect sensitive patient information.


Join us as we explore how data and AI are revolutionizing healthcare, from enhancing patient care to improving operational efficiency. Discover the challenges and opportunities in leveraging these technologies for better healthcare outcomes. Don't miss this insightful conversation with Ankush Khona on the future of data-driven healthcare.

Sponsor
The Evolving the Enterprise podcast is brought to you by SnapLogic, the world’s first generative integration platform for faster, easier digital transformation. Whether you are automating business processes, democratizing data, or delivering digital products and services, SnapLogic enables you to simplify your technology stack and take your enterprise further. Join the Generative Integration movement at snaplogic.com.

Additional Resources


The Future of Healthcare: Data Integration and AI Insights with Ankush Khona

Dayle Hall:
Hi, and welcome to this latest episode of Evolving the Enterprise where we're going to explore cutting-edge technology and thought leaders from around data, artificial intelligence, and everything generative that we're hearing a lot about. Today, we're thrilled to have Ankush Khona. He's a distinguished leader in the realm of data and AI. So exactly what we want for these kinds of podcasts.

He serves as the data and AI advisor at AuxoAI, where he's going to give us some details around that. They are at the forefront of integrating advanced data strategies in healthcare. So obviously, a real important industry taking advantage of data and see what's coming on AI. His insights into AI and data integration and the evolving role of chief data officers have been instrumental in shaping a lot of the modern healthcare solutions that we see. So without further ado, I'm going to introduce Ankush. Welcome to the podcast.

Ankush Khona:  
Yeah, Dayle, thank you. Thank you very much for having me to the show. It's an absolute pleasure to be here. I’m very excited today to share my insights and my experiences with the audiences.

Dayle Hall:
Yeah. Me, too.

Ankush Khona:
So let's dive right in.

Dayle Hall:  
That sounds good. Look, what I always try and do is give the audience a little bit of context. Tell me a little bit about your background, how you got into the data and AI realm, whether it’s something you studied and then grew from. And then also tell me a little bit about how you came to focus specifically on healthcare.

Ankush Khona:  
Yeah. It's a very interesting realization [to wrap] that I never envisioned in my life that I would be in health- and to be very honest with you, I think it just happened by the nature of how my career has progressed. Going back in 2003, I was given an opportunity to be part of one of the leading, I would say, global healthcare organizations, Novartis, its data center in Switzerland. Through that opportunity, I just continued to be part of the healthcare. And as I more and more got involved within the healthcare  domain, I just started to really appreciate the complexities of this domain.

And you look at the overall up-front manufacturers, to the health insurance, to the providers, to the wholesalers, to the pharmacy, to the consumer who was actually taking all these, it's an amazingly complex ecosystem. So that’s just [inaudible]. The healthcare never left me, as I say to everyone, and I just continue to become very deep into this entire domain. And I still get fascinated when, on a daily basis, I have to learn new things about this domain and it challenges you to the core.

Dayle Hall:  
Yeah, that's great. It's one of those areas, I think, specifically that whilst it can definitely benefit from new advancements in technology, we've already seen it in so many different areas in healthcare, but it's also such a- to be so cautious around patient data and privacy. So it's one of those unique- it’s a little bit like banking and finance with all that. Yes, if you're a technology company, you don't want to sign people up to Marketo emails that they don't receive. But it's not as critical to have that kind of safety and security in the technology. So I'm sure that brings a new level of challenge around healthcare.

Ankush Khona:  
Oh, absolutely. And you could actually get fined if you keep sending marketing emails to a patient who has told you not to send the emails. So again, there are these nuances where you don't get penalized for any other industry. Yeah, that has its own nuances.

Dayle Hall:  
Yeah, absolutely. All right. So let's dig in. We have a couple of different topics that we're going to go through. Let's start with data strategy, how you use data in and around this area of healthcare. In general, for people you advise or the organizations that you've been involved with, where do they start? How should a healthcare organization necessarily begin around actually creating a strategy that aligns to what they're trying to do in their business? Because, obviously, there's already so much data out there, but how does one go about or a healthcare organization go about creating a strategy around the data that they have?

Ankush Khona:  
An excellent question, Dayle. If you start with the very fundamentals of if you're in a healthcare organization, again, as I said, it's a very multi-complex ecosystem, so you really have to understand who you are. It starts with that very fundamental question, and it really needs to understand what your overarching vision is. The foundation of data strategy has to align very cleanly to your business strategy. If it doesn't, then no matter how much ever you do, it will never be able to meet the outcomes that the organization was looking for.

To me, I have started my journeys with all of these Fortune 500, 2000 clients, including health start-ups, I would say. The fundamental question I always ask is, okay, what is your vision, what is your strategy, and what do you want to be. That’s the fundamental question every organization needs to ask. The second and the foremost thing, and this comes down to not from a business standpoint but both from, I would say, a data and technology standpoint, who you want to be. Let me expand a bit on this.

So you have heard a lot of, I would say, words thrown out these days, like AI first, AI driven, data first, data enabled. And it's funny enough, but they are so different from each other. Being AI-first to an AI-enabled organization are completely different spectrums of skills and capabilities the organizations need. Yet, if, let's say, you just want AI to drive decision-making but you're never going to be an AI-first organization- for example, an AI-first organization is someone like Google, or an Amazon, or an Apple. They're going to be an AI-first organization because they're going to leverage AI for everything they do.

Most organizations, in my opinion, the healthcare organizations, are going to leverage AI, enable AI for effective or efficient decision-making, or embarking on these research and advancing the research within the healthcare domain. To me, that's the core of it. And you should not replicate these things because they're two different skills, in my opinion, and two different perspectives, in my opinion.

Dayle Hall:  
Do you find them, with the organizations that you talk to, are those terms a misnomer when they- because they hear about all the potential of AI and they say, we want to be AI first and everything. But I like what you said originally, which was the foundation of that strategy has to align with the business strategy. Do you use that initial question to then say, well, are you really AI first? Or you can use AI, but that doesn't mean you're going to use it in the same way as an AI-first organization?

Ankush Khona: 
I think it has to be. A lot of times, unfortunately, a lot of the leaders get confused because all these vendors and a lot of the organizations keep coming and selling to all these executives these fancy little things. But when you get to the very fundamental of first principles and first thinking of who you want to be and why you want to be, I think you will get your answers. And that's what I think- my request to my leaders has always been, please, take a look and think through this one because this is the most important question.

Well, it's not like you're setting in stone, right? You could go from one to the other as you continue to embark on this journey, but you have to start with some original premise. That's the depth. So to me, that's the broader question that organizations need to realistically answer. And then you get down to the fundamentals of, okay, let's say if I want to be an AI-enabled organization, what skills do I need? What talent do I have? What data do I need to enable? For what outcomes? Then every other aspect of defining the vision and defining the priorities, defining your talents, skills, understanding your talents, skills, where you are, and being very, very realistic about that, it's very important for you to realize the value of these investments that you are making.

I think most of the organizations just try to copy, say, oh, I want to be Amazon. But you realize Amazon took 15 years to become Amazon. It just did not become overnight. Do you have the patience and the perseverance for 15 years to become an Amazon? Do you have that long vision, long mindset to becoming that? If the answer is no, then don't fool yourself. Stay true to what you are and go after. So that would be my genuine suggestion.

Dayle Hall:
Again, I think that comes down to, look, a business strategy is usually on a longer timeframe. I think thinking about data and AI strategy on that same timeframe that aligns with that could at least have them make the right decisions up front and also set the right expectations of what they're going to get out of an AI investment, either early or later in the road map, let's say. So when you go in and talk to organizations- and everyone's talking about AI now. They're talking about generative capabilities, the GPT capabilities.

So when you get the opportunity to talk to organizations, healthcare organizations, how do you advise them on what timeframes to be thinking? Because some of the things that I've heard on previous podcasts with previous guests is their recommendation is, first of all, start with a business issue. Will AI solve it? Look to have quick wins. So solve something small first and then grow from there. How do you advise organizations that you deal with around timeframe once they've got a strategy that they think they can use AI to help?

Ankush Khona:  
I think that's a brilliant question. Let me break that question into two different thoughts. I agree on the [former thing], which is, obviously, understand a business problem, solve it, pilot it, prove it, and move forward step by step, take a smaller journey. But there is even a macro aspect to it. And if you look at healthcare, again, it's a very complex ecosystem, you're dealing with patient health. How much do we really understand human body is a big question. There's a lot of uncertainty and unknowns within the own realm of- we’ve come across much further in terms of understanding human biology and human body, but there's still a lot of things that have been not fully understood.

But with that mindset, healthcare, in my opinion, is an infinite mindset. You work with an infinite mindset. It's not constant. You’re not building it for a finite set of API meetings. And to me, data is similar to that. All I want to say to all the leaders out there is have an infinite mindset, meaning just because you have defined the data strategy and you have built the data platform, doesn't mean it's going to solve your problem once and for all. You will have to continuously evolve yourself as the majority of these technologies are evolving, as the paradigm shift in the healthcare is happening, and as the convergence of different trends are happening.

Eighteen months before, nobody knew Gen AI was such a big thing. Now everybody's talking about Gen AI. Maybe 12 months down the line, it could be quantum computing. Twenty-four months down the line, it could be blockchain in healthcare. I'm just giving a random example, but you get the gist of it. As these trends will fundamentally keep changing, the constant to everything is data. AI is never going to be successful without data. Precision medicine is never going to get successful without data. So data is the underlying catalyst or the dependency for everything. [inaudible] infinite mindset of science.

We will build something that is adaptable and emergent, and we know that it's going to continue to change as these technologies get evolved and as healthcare, as human understands, gets evolved over a period of time. And we will bring these together in solving the problems on a year-on-year basis. So you build that infinite mindset, but then you make it very, very prioritized to the business priorities for that year. That's how you enable that.

Dayle Hall:  
I love that expression, have an infinite mindset. A lot of the time, again, you can solve certain problems, but one of the questions I was going to ask is how do you make sure you're adaptable to new technologies? But I think you just answered that question, which is if you build flexibility in your infrastructure, then you can potentially- depending on how much of a seismic shift the technology is, then you can have that infinite mindset. But if a new company came to you, healthcare or otherwise, and said, okay, but what does that mean? What does a flexibility in my infrastructure- okay, I have an infinite mindset, that's exactly what I want, but how do I practically build that into my current organization to make sure that I can adopt new technologies?

Ankush Khona:
Absolutely. That's where you start with the broader architecture in terms of, okay, how do I bring the current capabilities to the best of my understanding and bring it together in a way where it is more plug-and-play and it's not very, very point-to-point-based architecture. So you start to make it a little bit more a plug-and-play architecture, more modular. Let me give you a great example about this. I'm sure people know this, but I’ll still iterate it.

One of the companies that I have been impressed with over a period of years is Uber. And let me give you why. Uber started with what, fundamentally started with nothing but a ride-hailing taxi service. What it did was they built the platform that can [inaudible] foundation door. And then eventually, what it did, it made it so modular. And it not only had the ride-hailing service, it now has a food delivery service. It's the same platform, by the way, think about it, just adding capabilities on top of that.

However, what it started was the health service, the Uber Health, and it started the Uber transportation. But if you look at the underlying platform, it's the same platform. And I think that's the mindset, that you build something that is so modular in nature. As your business moves around, you don't have to completely take out the entire platform, but you expand on it and you support what the business is looking for. So to me, that's why I gave the example of Uber. It’s a great example of how a platform could be built that is so modular in nature but supports the core businesses of how you evolve from a business standpoint. Did Uber do this probably 15 years back, I'm sure the answer is no. But they built it with a very distinct mindset.

And similarly, healthcare can take that same example, similar example. You don't have to copy or replicate it. Again, just goes back to the same theme. A lot of these healthcare organizations, a lot of them have so much of money or resources like Uber has, but you can apply the similar principles. And so, okay, I am a provider today, I would like to get into a health organization tomorrow, what does it mean for me? So I have to build a platform that allows that external collaboration with research partners, with technology partners, or whatever it is, allowing me to expand my broader services to my patients or my members in a most cohesive way. So you start with that broader platform vision.

Dayle Hall:  
And again, I think your example of Uber, obviously, they’ve proven that they built that flexibility. But it's good to know that even healthcare providers or people that are in the healthcare technology, they can still follow the blueprint.

Let me ask you a question specifically around patient privacy, regulatory compliance, governance around data. Because every technology, every company, every vertical has some level of privacy, but with healthcare, they've got to be significantly more controlled. Is there a different way of thinking about technologies or data or AI around healthcare that you advise people because of the nature of patient security and compliance data?

Ankush Khona:
Yeah, look, I'm now going to go back to the very basics. The fundamentals of patient privacy and data security, it’s the most, utmost- you have to apply the highest level of standards and the frameworks for protecting that information. However, to think about it, the patient in its nature, end of the day, they are dealing with health. And there is an emotional aspect to it. So you look at the financials, like a banking sector and health, there's a money part, there’s a health part. It's more emotionally attached to more people.

When you start to look at these AI capabilities to be built out, my guidance to the organizations have always been don't start with external focusing, start with internal productive improvements first and make sure that your AI models are tested enough to ensuring that it's providing the right results before you just float it out right to the outside world. If you're a retail company, let's say you just built a model that, for whatever reason, could go wrong. The maximum that could have happened is it could impact your financial statement for whatever reason. But it's not going to do beyond it. The damages, it's not going to be huge, unless you do something really damn stupid. But in case of health, if something goes even- when it's gone wrong, you're just talking about the health of a human being. So that is the subtle thing that the difference between a regular- like other industry was that health is one thing.

And then from a security and privacy perspective, think about it, the healthcare patient and member is evolving from being a patient to a consumer. And what I mean by that is that a lot of people, especially the newer generation, they want to control their own health. They want to manage their own health in a very proactive way. So it's starting to become more of a consumer-driven industry. So now, how do you ensure that you are opening up the data so that, let's say, a Dayle or an Ankush could share with other providers? Or let's say, just for the sake of argument, you go to Dubai and you felt sick, and you wanted to share your results with the provider, they should be able to because then, accordingly, he will be able to treat you really well.

Eventually, I think this health [pack] should be more flexible in a way where the consumer has the ability to share the information for the betterment of their own health.  But I think today, it's very, very, I would say, confined in a cage where- I think the reason why healthcare has not evolved much beyond a certain extent is the exchange of information between different parties. And today, it’s still a big challenge and a huge challenge for the industry to share the data because everybody wants to control the data. So to me, that's the biggest thing. I would love to see where it opens up a little bit more in a controlled fashion when the user has the ability to doing it, and they control what needs to be shared and what doesn’t need to be shared.

Dayle Hall:  
Yeah, for sure. So let's move on a little bit further to talk specifically around the AI or data integration in healthcare but beyond potentially managing someone's data. Are there other use cases you're seeing around data integration? Or even things like AI in hospitals where they're using speech-to-text? Are there technologies that they're using to save on manual data entry? What are some of the practical applications you're seeing? Practical applications, but who’s doing a good job? Who's really taken advantage of better access to data using AI and so on that you see when you talk to clients?

Ankush Khona:  
Yeah. Look, I think the majority of the healthcare, in my opinion, has always been there. They have always used the latest and the greatest of the technology. As I said, it's made available. I think it's just the integration of these capabilities to forming the experience, I think, is just a little bit more challenged. Let me give you a few examples.

I think the rise of precision therapies, so where there has been rapid advances in artificial intelligence, DNA, RNA sequencing, you may have heard about CRISPR, gene editing, lab automations, they have really spawned these new therapies, enabling the treatment of diseases previously considered not possible or intractable. So I think the rise of precision therapies, I would say, multi-omic technologies where your ability to really understand the protein structures of the body- the organizations are using extremely, extremely advanced AI capabilities to understanding these protein structures and everything. I'll give you an example.

There's a company, Insilico Medicine. In 2021, this company basically said they have identified a preclinical candidate- and just for the people who do not understand healthcare, preclinical is just before it enters the human trials. So it's a stage before the human trials. And they identified it through an AI platform. So they identified the molecule, they identified the targets, they went through the preclinical trials. And in span of 27 months, they went from preclinical to the first-in-human trial, which was a significant time series optimization from a traditional clinical drug development model.

Again, this is one of the successes from many failures that the industry has gone through. But slowly and steadily, the organizations are starting to really, really leverage artificial intelligence, extreme deep learning neural networks, to really accelerate the drug discovery. They are starting to look at operational efficiencies in terms of, the example that you gave, speech therapies. There was a recent, I would say, a survey done where most doctors and nurses within the United States are under a deep amount of pressure for the amount of work they are doing. So they're burnt out quite a bit.

And so there are a lot of these start-ups that are starting to look at AI as a capability to help them with clinical documentation, making sure that everything that they're speaking is getting converted to proper clinical notes, summarization is happening, charts are being created in the EHRs. So obviously, the healthcare organization, in my opinion, as I said earlier, I've always been there, but now they are adopting with full force.

Now I still feel the biggest challenge for all of these folks would be integration and exchange of information between different parties. And if they can figure that piece out- I think everybody wants to do it, but nobody wants to do it. I think that's the challenge with healthcare. And as long as they can do it, I think it will open up amazing amount of opportunities for the healthcare domain that has sense of helping really getting there faster.

Dayle Hall: 
I think, again, you mentioned it yourself, learning from past mistakes and being able to do better on clinical trials or speed that time up, I think that's important. It's hard to not be cautious when you're talking about speeding things up with patient data, for example. I get it, people don't want to make those mistakes. You mentioned, you used the words operational efficiency. I have a question on that because we all want technology that is better, faster, cheaper, right? We all want that. But are there things that you see within organizations, within healthcare, where they're focused on is it to get better or faster, or is it to cut costs. Where do healthcare organizations start from? Are they trying to cut costs? Are they trying to do better healthcare information for patients? Because, as I said, we all want all those three. But where do they start from around operational efficiencies?

Ankush Khona:  
It's an important question. I don't probably know the full answer to it, but I think every organization obviously starts with the very core fundamental of improving the patient outcomes, which is the health of a patient. At the same time, they want to reduce the cost of serving so that, obviously, healthcare becomes cheaper. Now is the second piece happened? The answer is absolutely not. In fact, it has probably become really, really bad in case of a lot of the disjointed capabilities in the organizations are leading to- you could see on social media where people are complaining about how they got scammed with this huge amount of bills and unwanted services that were offered.

So I do believe that the technology- for example, like when they moved into the COVID era, the organization was immediate to go into the telehealth, digital health capabilities to solving the problem versus seeing the doctors in the- so I think that's where the technology was used for the most, highest advantage. I do believe that a lot of organizations are starting to focus on Patient 360. They've been looking at different points of care across the entire journey and trying to figure out the effective treatment care management capabilities for the patients so that they can be treated rightly. They're starting to look at more predictive analytics capabilities to ensuring that they can ensure that readmissions are not happening, thereby they can reduce the cost.

So obviously, the organizations are doing, but again, if we look at in its entirety, is it happening for a patient? The answer is probably no. I think this is like 20 years back, the banking system was you go to a bank, you go to an ATM, you could go to different lengths of different touch points, the banking did amazing, great job of integrating the services better in a cohesive experience for the customer. I think healthcare really needs to do that, and [bending] these cohesive experiences end to end for the patient in a well-defined service.

A service by itself would be great. So you go to a bank, and your service is great, but let's say you went to the ATM and ATM did not pump out money, you would still say the bank service is bad. So all I'm saying is the service by itself may solve the problem, but it will still be considered as a bank service if it's not solid in its entirety. And so to me, the healthcare is at that stage where the entirety of its services is not being handled. It's being handled in piecemeal solutions. And it's trying to solve that physical capability, but it's really not helping the member as a whole because they're getting hit by something that they were not expecting.

Dayle Hall:  
Yeah. And I think that comes back to what you said originally, which is if a business strategy is to provide, let’s say, immediate and better patient care, then you're not deciding whether it's better, faster, or cheaper. It should be all of the above.

Ankush Khona:
It should be all.

Dayle Hall:
That's right. So you shouldn't be making these decisions around data, around using AI, just to solve one thing. You should be looking at it holistically, which I think is-

Ankush Khona:  
It’s an end-to-end experience of a patient, and how do you make sure it is built to ensure that you are truly helping that patient to be proactively managing their health, being healthy, but also letting them know what's the price and the cost. The cost transparency is the biggest problem statement today, where most of the consumers are getting burdened out by healthcare, purely because they really do not understand what is the cost of healthcare.

If I go and walk into a hospital for an x-ray, there have been instances where people have called the health insurers, got an exact amount, but when they get the bill, it's a completely different amount. They cannot challenge it because- you could challenge it, but it's not going to go anywhere. We all know that. And it's like a mafia system where you just end up paying it because you’re just not getting the service and you have to pay for it.

Dayle Hall:  
Sounds like me when I'm trying to claim on my car insurance, but that's a totally different topic. You mentioned just something previously about you think integration of data is one of the key challenges. People want it but a little bit reticent. Can you explain a little bit more about that? How are we going to get to a point where organizations feel comfortable enough that they can manage all the integration points, the different types of data, so they can actually truly deliver something like Patient 360? How are we going to get to that point where people feel comfortable enough that we can solve the integration challenges?

Ankush Khona:  
I think large part of it is starting to get resolved through the intervention of the governments. If we look at outside of the United States, a lot of these governments are actually starting to get into integration of data. The pandemic has forced them to think through. Now I think in the United States, I think the CMS, recently, two years or three years, it passed the interoperability ruling that is now forcing the payers and the providers to exchange information. Now is it happening at a full scale? The answer is probably not, but it's a step in the right direction. So I do believe that in the coming few years, I do think all these organizations and students start to realize that without data, they really can't do much.


Interoperability between these partners will significantly improve. I think we are already seeing the reduction of manual data entries from the smaller provider groups and even everyone. So they are starting to adopt these digital technologies that is forcing them to reduce the manual data entries at multiple points, which will improve the overall data. And also, I think there are these start-ups that are doing the job for integrating their data. So they are going and building these partnerships and collaborations for helping to bringing the data into one place.

Now last but not the least, I would say, I think this is where technology can really help. The rise of artificial intelligence, Gen AI, [innowebs] would lead to organizations creating synthetic datasets that will allow them to really understand- I mean, it might not be a real-world thing, but enough close to a real-world data that would allow them to start building the right capabilities for the patients and the members. That hopefully should reduce down the nature of dependencies on organization. But I think that's, to me, I do believe that those are some of the areas that we'll get into in the next coming years.

Dayle Hall:  
Yeah. You mentioned, obviously, LLM and Gen AI capabilities. I think one of the things that- as an organization, we're seeing a lot of interest in the capabilities, not just to handle the integration part but enabling organizations to use the LLMs internally, helping them build their own LLMs in conversational interfaces. But there's this reticence to take an outside technology thinking, well, do I lose my data? Or is my data going to end up training an external model that I don't want to happen?

There's still a lot to figure out. It's exciting. But our third section of this podcast, the thing I wanted to touch on because you have good experience here, is all the things we've talked about and the challenges. This generally falls in the realm of different roles within the organizations. There's the rise of the chief data officer as a role, and a lot of it kind of sits with them. Even if they're not necessarily running all the pieces of integration or they're not managing all the AI, generally, they are the people who are responsible to make sure that the data is feeding these models. So how does this role in healthcare- is it different in healthcare versus other organizations? Are there specific things they have to care about? And how does a chief data officer these days cope with the massive requests or technologies that they have to implement?

Ankush Khona:  
I think there's a bunch of questions in that one question, so let me break it down. How does the role of CDO differ from other industries to what's unique to healthcare? I do think that's a little bit unique. I still would compare healthcare to finance where there is bunch of 15 different things that are collaborative healthcare. So like under finance, similarly, healthcare is like, you could be a payer, you could be provider, you could be a pharma manufacturing, you could be a med device company. So to me, I think depending upon where you play, the role could be subtly different. And then also depends on if you are, let's say, a pharma manufacturer, you are a global organization, so you have US-specific rules, Europe-specific rules, Japan-specific rules.

And so the chief data officer, if it's a global chief data officer, obviously, they have to ensure that they have a good understanding of these regulations within those countries. So it's one subtle difference. I think if you are a provider like a hospital operations, typically, what I have seen is a doctor with a very deep clinical background is typically appointed as the chief data officer because, obviously, it's all about human health and clinical understanding. But if you are a pharma manufacturer, you generally are looking for more strategic leader who has the ability to think through more broader than just clinical. So you have commercial operations, regulatory affairs, so a lot of these other areas that, combined, are covered by pharma manufacturer.

So to me, I think depending upon where you play within the healthcare, the role and the qualification of the CDO could vary significantly. And so that's at least one thing that was subtly different. I think the second question was mostly around-

Dayle Hall:  
How do they cope with the influx of requests, technologies, and not become a bottleneck, essentially?

Ankush Khona:  
Unfortunately, whether we like it or not, the reality is the role of the CDO, unfortunately, is not very well defined within the industry. And let me expand a bit on what I mean by that. Typically, the CDOs are expected to come and build these data platforms at a much faster pace and start to generate value out of it. What the organizations fail to understand is the integration of the data within the healthcare is a very, very complex process, and it takes a little bit more time than you would generally do it, because even though there are standards, most organizations don't use those standards. So like, if you are in banking or finance, you're using these standards that are easy to build capabilities out, which is not the traditional case in the healthcare, even though there are standards existing. So that's the other piece.

Typically, also, everybody wants to build or focus on data. But in all fairness, people do not understand or are willing to understand the underlying complexity of what it takes to build a data. Everybody wants to have the value generated. And it's not going to happen. That’s the reality of the situation. So to your question, what happens when they’re getting [inaudible], most of the CDOs, they do get [inaudible] insane amount of requests to solving the problems.

I think there have been organizations who have- again, it comes back to the initial thinking that you and I had, a longer-term thinking, how do you build something for a longer term but solving smaller-term problems. And also, I think if the business and the key executive leadership understand that the CDOs are here to not create value, they are here to enable value generation. So what I mean by that, they themselves, by themselves, will not be able to create value. They will enable value by enabling these capabilities for the organization. And that's how you should measure them.
So what you were doing yesterday, what you're doing today, have you improved upon? The mandate should be a little bit more different. You're not building something for today. You're building something for next three to five years that you can then start to realize the value of it. Obviously, people just do not have that amount of timeframe and vision. And that's why most of the CDOs have a very shorter duration within the industry, which is probably like two to two-and-a-half years. If I'm not mistaken, that's the latest. And then there are other reasons for it. They sometimes get confused where the chief digital officer- sometimes a role is divided with the chief information officer.

Again, I go back to my first statement is, do you want to truly become a data-enabled organization? Then the role of CDO is very important, and you need to craft the right responsibilities for the CDO role. And the CDO needs to partner in the right way with the CIO role in ensuring that they both are working towards that common overarching vision of becoming a data-enabled organization, I think most of the case does not happen.

Dayle Hall:  
Yeah, look, I think you’ve really hit the nail on the head there. I used to joke- well, it's not a joke. I know CMOs have a short tenure in organizations. It's between 18 and 24 months. And I read recently, you just brought it up yourself, chief data officers is about two years as well. But what I think you're saying is if you can set the right expectations that you're here to enable value generation, not deliver all the value, then do you think that would help the organization understand what a CDO is there to do? And do you think that would help from a tenure perspective for them?

Ankush Khona:  
Yeah, it would help, definitely. And then the other thing, my advice to a lot of CDOs is also just don't go and start to build everything from scratch. That's the other mistake- a lot of people end up doing it. And I'll give you an example of a company that I, in my initial research, came across. This, fortunately, is an Indian company, Asian Paints [inaudible] was paint and chemicals and everything. They implemented a platform back in ‘70s that started to bake that data in. And it's been 50 years, that company is probably one of those very few companies in the world that have generated profits at 30% year-on-year for last 50 years. And the reason they have done that is just because that initial vision set in ‘70s or 80s, they just did not go and remove the platform. They continued to enhance the platform. Every new leadership came, they just did not throw it out. They built on top of it.

I think the challenge today for most of the organizations is when a new leadership starts, they start from scratch. They outdo the work that was done previously, and that takes away a lot of good work that was done to generate value quickly. Again, unless it's a real, completely poorly built-out capability, that's a different story. But my suggestion would be always look what you can take, take that acceleration to the value while building the broader platform, and continue to build upon it. Same like Uber. Think like Uber, think like Airbnbs of the world, you have to really rebuild everything but continue to enhance it over a period of time as you continue to mature your data journey and your data capabilities. I think that's the one thing I would say that organizations can do differently.

Dayle Hall:  
That's interesting. I guess my last question around this kind of role, and again, I think this could apply to healthcare or to any industry for the chief data officer, if they think about enabling value generation and the things that are going to- maybe they don't start from scratch, but how should they be held accountable to- what are the metrics that they would then be held accountable to, so the organization, the executive team, the people who are investing in them, they understand? What is a measure of success? How should they be held accountable so it's clearly understood, and so they're not necessarily worried like, Oh, my God, unless I deliver X revenue or whatever in the next 12 months, I'm going to lose my job? What's a good measure of success?

Ankush Khona: 
I think there are quite a bit of metrics that organizations can employ, and it starts from the baseline of where you are today. For example, how much time does it take for you to solve a data request? Let's say it was three days before. Did you bring it down to one day? Great. So that's the measure of success. It’s a very quantifiable number.

The quality of data, the accessibility of data. You go to any organization today, and I promise you this is going to be the first time, everybody's going to tell you the same thing, I can't access my data. And you deep dive further and say, okay, how much time does it take for you to access the data? Can you solve that? These are the low-hanging fruits that the CDOs can go after. They don’t have to solve the broader platform problem. Yes, the broader platform problem is obviously critical to the success for the long-term story, but there are these very small low-hanging fruits that can be solved for.

Improving the literacy program of data within the organization is another good metric. People make card-based decisions. Can they start to make data-driven decisions? It could be in Excel, or whatever it's worth. Then they look at data but make a decision. You are forcing them to start thinking data-driven key mindset.

So improving the data quality, accessibility to data, I would say, the literacy of the data in the organization. If it’s a healthcare organization, how are we building something that is improving the quality of care? Is it improving the population health? Is it improving the member satisfaction and engagement? Those are very, very core metrics that we can go after.

Dayle Hall:  
Those are good ones. I know we could talk for a lot longer, but my final question for today, of all the things- you're obviously very aware of a lot of the data and integration capabilities and AI, whether it's AI-first or AI-enabled capabilities, is there something specifically that you're excited about to see with some of these technologies? In healthcare or just in general, what are you really looking forward to seeing or what are you excited about? What do you think something- a capability that's coming that you think could be a game changer? What makes you excited when you think about some of these new technologies that are coming?

Ankush Khona:  
I think, to me, as of today, it's the artificial intelligence and how do the leaders cope up with this new trend that is changing on a daily basis. It's even driving me crazy as an individual who likes to read a lot, but I'm still not able to even get in front of the daily advancements. So I think it would be really good to see how enterprises adapt it. I'm really, really interested in trying to see how do you adapt when it's changing so fast. And I'm really looking to talking to a lot of people seeking their collaboration and partnership and really trying to understand what's the best way to do productionize it.

So to me, that's the biggest challenge. I believe, as of today, the enterprises are getting into it. And that's something to [inaudible] next one to two years. And maybe after two years, there could be something different, but for now I think in the next 12 to 24 months, I do believe that's the biggest area of fascination for me. And I want to go after and see how do you productionize [the sort of scale] that organizations can truly, truly deliver over and starting to look at it.

Dayle Hall:  
It was great. Ankush, I really appreciate your time. You clearly have a lot of background. I love the fact you have excitement about it too and makes me- as a patient at some point, it makes me feel better, there are people like you helping these healthcare companies get better. So thank you so much for being part of the podcast today.

Ankush Khona:  
Thank you. I really appreciate for inviting me, and I'm looking forward to many more discussions like that.

Dayle Hall:  
I hope so. Thank you, everyone, for joining us on this podcast episode, and we'll see you on the next one.