AI Revolution Will Come Down To Who Has The Most Differentiating Data And The Highest Quality Of Data

Welcome to Season 5 of The Big Unlock. I’m Rohit Mahajan, Managing Partner and C.E.O. at BigRio and now, Damo. This season carries forward Paddy’s legacy and joining me for an exciting discussion, in this episode, is Vineela Yanamreddy, Chief Information Officer for United Medical Center also known as Not-For-Profit Hospital Corporation in Washington, D.C.

Q: Welcome to Season 5 of The Big Unlock. I’m Rohit Mahajan, Managing Partner and CEO at BigRio and now, Damo. This season carries forward Paddy’s legacy, and joining me for an exciting discussion in this episode is Shahidul Manan, Chief Data Officer for Bon Secours Mercy Health, Cincinnati, Ohio. Tell us a little about yourself and your organization.

Shahidul: I’m Shahidul Manan, the Chief Data Officer at Bon Secours Mercy Health and Vice President, Digital Health at Nordic Partners. I’ve been in this space for almost 20+ years now— in financial, for over a decade, then in high-tech software space running analytic software, and finally, in healthcare for the last seven to eight years. It’s an exciting time to be in technology, especially given emerging innovations in data and AI. 

Currently, Bon Secours Mercy Health is an enterprise of about 70 hospitals—a hospital system and six subsidiaries—and about 70,000+ employees and 20 million patient lives. We’re in seven states and through an acquisition, in Ireland, as well. As the CDO, I manage pretty much anything and everything that touches data, including data platforms, the cloud, digital innovation, and focus on analytics while serving all the systems, data productization, and AI innovation. I’m also focused on our platform and data AI product commercialization to some extent. Commercialization entails sharing the goodies with other systems and in the healthcare space, so that our innovations can actually touch other systems and their patients while garnering revenue for us. 

Overall, healthcare, right now, is a great place to be in because of the innovation and all the direct impact that we can have on patients and our healthcare services. Historically, healthcare was or has lagged a little in terms of digital transformation. Our first true digital transformation began with the ACA—Affordable Care Act—where the EHR became almost like a mandate. The digital platform—the EHRs—started capturing all digital footprints for our patients and providers and everything we did to create the data that we needed. In the next phase, we will use that data to drive everything we can from building insights, data-driven decisions and operations, and data analytics-driven innovations for patient outcomes, quality of services, and various patient engagements. 

Q: Prior to Bon Secours Mercy Health you were at Mass General in Boston. Tell us what got you excited about getting into the healthcare industry? This was not where you started originally but this is where you are spending most of your time now.  

Shahidul: Absolutely. I was with Mass General before and in a similar role as now. I drove many innovations with data AI there, as well. Today, healthcare is becoming more and more interesting for two reasons.  

One, it has been a little bit of a greenfield compared to other industries as the focus is more on digitization. There are tons of greenfield innovation opportunities that are emerging, today.  

Two, healthcare is likely to show highest growth in data sets. The AI revolution we’re in is eventually going to come down to who has the most differentiating data and who has the highest quality of data and healthcare, I think, has an edge in this matter— having the highest growth and volume of data. That also places on us the burden of harvesting that data and enabling it. There are plenty of challenges to this, starting with the fact that 80 percent of healthcare data is actually unstructured data in the form of physicians’ nodes or X-ray images. Therefore, building the truth, getting insights, or even building predictive power from these, is more challenging than ever. In terms of challenges and opportunities, this is a great place to be in.  

Lastly, this is very much a mission-driven effort and a mission-driven space, which also motivates me for I know that anything that we do or our team does in terms of innovation actually touches human lives and makes a difference in patient outcomes and the cost of healthcare—that we, as a nation, and actually globally, are struggling with. That’s why I get excited about health care and find it extremely rewarding.  

Q: You spoke about the explosion of data, especially in healthcare and I couldn’t agree more. We are seeing this increasingly with remote patient monitoring and all the devices that help the patient’s journey. How do you think about two things—building a robust data engineering and data infrastructure and leveraging these and AI to generate use cases? 

Shahidul: This is a pretty complex one in terms of how we are going to tackle this big challenge. My strategy is—and that’s what I would suggest to the others—to think of large-scale enablement, first. How are you going to collate all your data in an efficient, consolidated, aggregated manner such that you can leverage high quality or better processes to technically enable your enterprise to generate all kinds of use cases?  

No longer are we in an era of daily and weekly reporting. We’re getting into real-time analytics and insights to build that infrastructure, platform, and enable data aggregation with proper quality measurements. With quality oversight, proper data dictionary and governance, and AI comes a larger responsibility of using data governance for AI governance. For all these to come into play, there has to be some enablement. That’s what I call the foundation. Anyone, thinking about this space then, needs to start with that.  

The foundation needs to be bigger than what your EHRs offer. It needs to be a more independent and innovative workspace with these various bells and whistles of governance and innovation tools so that it can move from real-time analytics to AI-driven applications, productization, and building workflows so physicians, nurses, and hospital administrators can use it. The more you operationalize it, the better benefit you get. Even on the Gen AI front, it’s no longer a science project of some kind where you try it on an experimentation level for that’s just the beginning. You want to think about how to operationalize it and actually put it in production—in the hands of the field players—so you can actually see tangible value coming out of it. Therefore, my suggestion is start with that foundation—think big, but start small.  

In the next phase, look at harvesting various types of value use cases. You want to pick the use cases that actually provide that tangible value and start with all these field stakeholders so you get a pragmatic use case you can build, validate, and then, actually put to use because the value comes out only when you use it. Keep that in mind, start the innovation there, and put these in the hands of the field level players—physicians to hospital administrators—and that’s exactly what I’d call innovation space of data and AI.  

The final one is never ending continuous improvement geared toward value extraction. This will enhance operational maturity levels of your data and AI. When you are at a stage where you have built your foundation and proven your value through various high value use cases, you can aim at operationally efficient levels and make it a part of your business or enterprise strategy. Your strategy and your AI initiatives cannot be separate—all of this needs to be at a space where systems, healthcare providers, payers, or any stakeholder in healthcare, when they build their next level of strategy for service, or product, or improve what they do, need to think of the additional value services or products they can provide with AI.  

Therefore, it needs to be embedded with their strategy and they need to feel confident that they have the capability and the value use case or value proposition to connect them to and build their next level of strategy together. That’s how I would approach this.  

Q: You’ve mentioned strategy and innovation. You must be participating in many C-suite meetings at your organization and across the healthcare segment. What are the business problems that you hear of from stakeholders and senior management teams? How are chatGPT and Gen AI influencing these?  

Shahidul: Everyone’s excited about Gen AI and its potential but in many cases, they don’t know where to start. In many cases, they’re worried that they will be left behind. So, it’s an interesting time actually for the C-suite to be challenged with this opportunity presented by AI. What I hear most from my colleagues and the leadership that I work with is that it is several fold. 

One, everyone’s worried about the enablement. Many are wondering if they are ready for undertaking this type of innovation or even taking advantage of Gen AI, machine learning, or even advanced analytics type of activities. Honestly, most of them feel they’re not ready. I don’t want to put an exact number to this, but it would easily be between 70 to 85 percent that think they’re not fully there with their data technology and AI capabilities to even explore to the extent that the right value can be extracted for their enterprise.  

Two, the interesting part is that 90 percent or more—and this is also from various studies—feel compelled to do something with this. They are committed to some extent with funding or with their strategy to do more with data AI from platform to innovation. So that’s the good news.  

Three, I have also been observing aspects to do with governance and rightly so, because we’re in a highly regulated environment. Being a part of the healthcare domain means we are the custodians of patient data so we have to be extremely careful not just from the regulatory perspective, but also from the ethical and clinical usage standpoints of what we do. All of this needs to be orchestrated with the strategy and then, only must tactical approaches be undertaken as many healthcare organizations are doing now.  

Lastly, on the governance front, I must add that we will see a huge change in the next six to 18 months in the regulatory landscape. The AI Act is already in place and when it gets executed, it will start to demonstrate new opportunities and new challenges for businesses and organizations such as ours in healthcare. The White House has set up various task forces to come up with new regulations for managing AI better and these will start hitting the ground in this time horizon. As technologists, practitioners, and leaders in the innovation space, we must comply with these regulations and leverage the opportunity while ensuring we are prepared and have fortified our organizations to manage it appropriately. 

I think that the next six to 18 months will be more like a warmup. We will need to see how we can better prepare ourselves and enhance our capabilities to manage the regulatory environment better for what we want to do and what we see as the opportunities with Gen AI in the next couple of years.  

That said, we already are tinkering with a lot of Gen AI type innovative use cases. Many organizations are actually trying these out, which is a good thing. It means, sometimes you have to get your feet wet just to understand where things are. It’s at that stage. However, with these maturing, there’ll be—even though it may sound overestimated in some cases—some slowness in the healthcare innovation space. But in the next three to five years, there may be a tsunami of products and services in this space.  

Q: With regard to regulatory oversight, how do you innovate in your space? How do you encourage your team and yourself?  

Shahidul: Besides the technology hat, we have to also wear the leadership hat. In addition, our evangelism hat has to help build that influence with the senior leadership and drive that influence to motivate our organizations.  

Starting with the evangelism, I have in place major programs to educate and train across the enterprise so that everyone gets on board and understands what this is and how it can be utilized. We focus on democratizing the data and AI capabilities so everyone can try them out, get onto the field, and benefit from access. That is also an enablement that we constantly work on—how to democratize data and give it to the data scientists, any department, or any business executive so they can see the data and make data-driven decisions.  

Another big part is about building the strategy, showing the value incrementally, and convincing the leadership about the value proposition of this whole digitization with AI. All these need to be orchestrated and so I also work closely with the team to help them comprehend the vision, understand the value proposition, and get the inputs and feedback so they feel excited about what we work with. It has to come from all sides—top down, bottom up, and sideways. We have to talk to the stakeholders, understand their pain points and how we can help them better. So, an upscaling of the organization is another focus area because the technology is exciting and constantly changing. It’s important to ensure that the team is excited and focused on constant learning—a state we are in since everything is coming together.  

Lastly, I have actually established a pretty robust governance model, even though I think there’s more to work on and do in terms of maturity. I work very closely and am lock in step with my Chief Privacy Officer, the Chief Security Officer, the legal counsel, and the IT teams so everyone feels they have a voice in this. We need everyone’s voice because it’s complicated and we are in a highly regulated environment. It is important to balance the compliance with innovation and the best way to do that is build that partnership, education, and understanding across the board. 

Q: We have launched a new offering recently in the market— it’s an Gen AI workshop that we have actually conducted at one of our client locations in Houston and given the positive response, we conducted another one at the Harvard Club of Boston. That’s our way of bringing some upskilling and education opportunities to our clients as well. In that context, can you elaborate on the data platform that you have built and are now commercializing? What does that product offering do? How can it benefit other healthcare systems and perhaps digital health startups?  

Shahidul: Thank you for bringing that up and for evangelizing on Gen AI. I’d love to hear more about the workshop because that’s exactly what we need. We need those words out and get everyone to understand what this means, the challenges it entails, the opportunities involved, and how one can get started, better. I certainly look forward to that.  

In terms of our platform, we have taken it to the next level by making it multi-channel capable and building into it greater de-identification and anonymization capabilities. Those are our custodian responsibilities—to make sure we are, as guardians, doing everything right to protect the data. 

That said, we are also looking at opportunities and partnering others—pharma companies to data vendors—to look for opportunities where we can help them innovate and share this anonymized or de-identified data for greater good while garnering revenue for the system. We almost have such a partnership with six vendors and partners. Not only is that garnering revenue for us but we want to build it as a more robust platform. We have more interested parties approaching us from across startups and pharma companies and system organizations because, I think, the more we can share data in this way, the higher our chance to become a larger ecosystem that can help solve the larger healthcare problems.  

Sharing innovations will also help those who do not have the budget or the capability to do everything in-house. For us as a system, we want to diversify our revenue and become a revenue center instead of a cost center.  

Lastly, we are also looking at several AI products that we have built and proven in-house from general health to COPD, patient risk stratification, hospital stay prediction or various readmission prediction models, all of which can help drive value-based care. We can certainly help bring those to market. 

The goal is to start looking at opportunities to bring to the market, provide that product and service to others, and garner more revenue. That’s a very, very strategic step and definitely, we’d be interested in learning more from Damo Consulting. 

Q: Would you like to peek into the future and tell us what is coming our way in terms of some disruptions and changes?  

Shahidul: I don’t have a crystal ball, but looking at the trends and the exciting movements, it’s a good time to be in data AI, healthcare, and technology as a whole. The next few years will be more exciting because we are going to see the outcomes of all the initial innovations that we have seen with Gen AI, LLMs, machine learning, natural language processing, data platforms, and Cloud driven analytics capabilities. There’s a plethora of technology capabilities that we have set up and are innovating with, today. So, stay tuned.  

I can see that personalized medicine, for example, will become big and that will actually make patients’ lives better. It will actually impact our day-to-day lives. I can see that healthcare cost and efficiencies will become more and more robust with the predictive power of data. That will help us manage our healthcare costs, better. We do need systemic improvement to solve the entire problem, but these are going to help tremendously. We are the biggest spenders on healthcare, but our return is unarguably one of the median or lower than median. We need to solve for that and this is a great innovation space that I believe is going to have significant impact; tangible and real impact in the coming days. I’m looking forward to it and hope you also get to test it, see the benefits, and participate actively in it.