Good morning, everyone. Welcome to day three of the JP Morgan Healthcare Conference. My name is Anne Samuel, and I'm the healthcare technology and distribution analyst here at JP Morgan. We're really excited to have Veradigm here with us this morning. With us, our Interim CEO, Dr. Yin Ho, and Interim CFO, Lee Westerfield, and we're also gonna be joined by the CEO of ScienceIO, Will Manidis, for a conversation about AI. With that, let me turn it over to Dr. Ho.
Well, good morning, everyone. Thank you for joining us in person and via webcast. I'm Dr. Yin Ho. I'm the CEO of Veradigm, and it is truly a pleasure to be here today. Thank you, Annie, for giving us the opportunity to introduce ourselves. We really look forward to beginning a dialogue with all of you about our company, Veradigm, and the opportunities we have to shape the future of healthcare. This presentation contains forward-looking statements and a disclaimer that you can read in more detail in our 8-K filing and on our website. So, here we go. Now I know we have not had a chance to speak in over a year, but we are truly excited to be here today to reintroduce Veradigm, provide an update on our financial guidance, as well as take a look at our business.
Veradigm is a truly differentiated health technology and analytics company. With longtime roots in supporting physician practices, to today, supporting physician providers, payers, and life sciences companies. We will be giving you a perspective of both our current business and of our future. First, let me introduce you to the members of the executive team speaking today, myself and Lee Westerfield, our interim CFO. Our team has the vision and the right expertise to accelerate growth in the next generation of healthcare tech. Now, I have been in the health tech and life science industry for almost 25 years. I'm trained as an emergency medicine physician, but I've built an entire career creating and developing businesses in data analytics and healthcare within life sciences.
I started in health tech as an e-health executive at Pfizer and later became the founder and CEO of several startups in data analytics and health information technology. I have led product innovation and transformation in this space, and it is truly exciting to be here today. Now, Lee and I, we have actually worked together before. In fact, we worked together at Aetion, a real-world evidence company, and I immediately knew he was the right person as Veradigm's Interim CFO. Lee is an experienced technology CFO who has grown multiple healthcare, health tech, and SaaS companies. He has expertise and a track record in building robust financial control environments, reporting timely and useful financials, and leading companies to profitable growth.
He has also spent 15 years among all of you as a sell-side analyst covering internet and media back when it was still fresh, and he will talk today about our very strong financial position. With that, I'm going to turn it over to Lee.
Annie, thank you very much again, and Yin, thank you very much, and I'm glad to be with you all. It's been quite a while since Veradigm has been in front of you, and so what I'd like to do today, after an extended absence of the company, is to reacquaint you with our financials and basic profile. Over the months ahead, we'll certainly be coming back to you on multiple occasions with additional and more detailed financials. Today, we summarize. So there are two topics: the audit and filings update and a financial summary. First things first, I'm sure you're all wondering where we are in our audit process, and you'll recall that NASDAQ granted us an extension to become compliant with its listing rules.
Also, as disclosed in our 8-K this morning, we received another expected notice from NASDAQ that we are not in compliance with the rule to hold a shareholder meeting each year, again, in the 8-K. We continue to work diligently towards our plan to become compliant with NASDAQ listing rules, and we'll make the required filings as soon as possible. In the coming days, we will be communicating to NASDAQ our progress towards becoming compliant with the filing rules and our plan for holding shareholder meeting, which we will hold as soon as we can after our requisite filings have been made. This morning, we also disclosed two other matters, call your attention. Namely, Veradigm has been named in a defendant class action suit, and we're cooperating with the SEC investigation relating to our audit committee's investigation.
Our policy is not to comment on ongoing litigation or investigations, so I'll not be saying more at this time on those matters. I would like to highlight, however, that with respect to our audit committee's investigation, as of today, we believe that any adjustments to our GAAP financials will be limited to non-cash items relating to the timing of certain impairments and accruals. Repeat, non-cash items relating to the timing of certain impairments and accruals. We hope that helps provide some clarity with regard to our financial impact of the investigation. Finally, with respect to our revenue restatement, we continue to estimate the impact will be a reduction in revenue from continuing operations of approximately $20 million in aggregate over the three-year period from 2020 to 2022.
I repeat, continue to estimate the revenue impact will be $20 million over the three-year period. On to 2023 numbers. About our financials. So I joined the company about a month ago, almost precisely, together with Yin, and tasked myself and with to perform a sober health assessment, if you will, of our company and its status, financial health, before I'd commit to financial guidance. So at that time, we were silent on a matter of guidance, and I'm here today standing before you with reset and refreshed estimations. We issued a summary of estimates this morning. You've not seen financials for over a year, so I'm gonna walk you through those estimates momentarily. First, I'd like to share my health assessment with you.
The state of Veradigm is fundamentally healthy, and that after a review over the past month by me. Our financials rest on a foundation that is fundamentally solid, and that foundation, you'll see, is evident in three factors: one, stable revenue base. Two, high-quality mix of recurring revenue, subscription revenue, as well as consistent profitability. And finally, the bedrock, net cash, not net debt, net cash on our balance sheet. So from my eyes, that sets a solid foundation, net cash and sustained profitability. The absence of a delayed filings and audit, sobering, in my judgment, has obscured attention from those healthy fundamentals, and understandably so. You haven't seen numbers for quite some time.
Our balance sheet strength, I believe, is demonstrates our fiscal strength, but also enables us to consider a wide-ranging set of alternatives for use of cash, as the board undertakes the finalization of budget, growth initiatives, product enhancements for margin, as well as returning value to you all as shareholders in share purchases when permitted. So bottom line on that point, we have the balance sheet integrity to be able to elevate ROI and total enterprise value when we move past the filing and audit period. About revenue, so today, you'll have read, we refreshed our estimated range on revenue to a range between $608 million and $622 million. About that revenue range, overarching, my intention is to provide clear and consistent communication about our performance with you going forward.
This estimated range is hopefully going to be clearer and more revealing to you, but I would encourage you to read our 8-K because there are specifics relating to elements in the GAAP revenue estimated range. This GAAP revenue estimated range includes GAAP items that are relating to customer litigation, which was favorable to us and included in our GAAP revenue. I take a quick step back from the canvas from a period of time in my life when I sat on the other side of the table and with you all. I call attention to three characteristics of our revenue. One, our provider segment is stable, consisting of a majority of subscription, highly high-quality subscription revenue, recurring revenue. Payer Life Science is growing and elevating its margin.
Adjusted EBITDA, again, health assessment, solidly profitable in the range of a little north of 20%, EBITDA margins. I'd call attention to your eyes to call attention to, in the news, our 8-K and our press release, we refreshed that adjusted EBITDA guidance, as well as the non-GAAP EPS guidance. I wanna make a point of clarity here. The adjusted EBITDA guidance of $122 million-$135 million, I should shed light on a detail. Different than the company's prior guidance in September, our estimates and our method of presenting adjusted EBITDA now excludes, previously included, now excludes the customer litigation settlements.
As I went about the past month and assessed how do we present information, one of the judgments that I made is that our Adjusted EBITDA numbers should be cleaned of wavering figures in the area of customer litigation settlements, pro or con. That explains a substantial portion of the variance. There are additional expense accruals, and again, I encourage you to read the 8-K for a little more details. And the takeaway, in my judgment, again, attractively profitable on the EBITDA margin. Additionally, we are free cash flow positive before acquisitions, things of that nature. Finally, net cash. Let me describe it quickly and suggest what we do with some of that cash.
Our net cash stands above $232 million as of December 31st, with cash slightly north of $440 million, and convertible debt outstanding of $208 million, and an undrawn facility. So what do we do with a clean balance sheet, and how do we assess choices? We have a range of opportunities, I and the board and our leadership team are assessing, and all are on the table. Return value to you as shareholders with share repurchase, which we have done in the past, are suspended from doing at the moment. Could should consider strongly as we regain listing requirements. Additionally, invest in the budgeting process we're undertaking now. Consider investing internally in product enhancements in areas that include cloud migration.
And finally, finally, and crucially, strategic investment in partnerships that can enhance our growth and modernize our product mix for a scientifically advanced future. Do all of that with an eye to appraising choices based upon ROI and enterprise value. So wrapping up, again, takeaways: financial health, stable growth, high mix of recurring revenue, consistent profitability, and a capital foundation, solid capital foundation, sitting on net cash. With making good use of that cash. Yin Ho has words. Please.
Thank you, Lee. So all right, I get to have some fun and talk about the business. As part of our reintroduction today, I will give you a brief overview of our business and the uniqueness of our capabilities that are driving current growth. I want to be clear that we are operating from a position of strength, differentiated assets, and access that makes our outlook very exciting, which is leading to what is the next generation of health platforms with health intelligence products. And we will cover in more detail shortly, but I want to preview why Veradigm is so well positioned in this moment. We have scalable, high-quality data assets. We sit at the intersection of providers, payers, and life science companies.
We have deep healthcare ecosystem expertise, and now with AI, we have the ability to jump forward with our product offerings by harvesting and monetizing the data and analytics of our existing business. Our business has evolved over the last decade. Veradigm divested its hospital business and refocused the company centered on our provider offerings, as well as two very important fast-growing customer segments: payers and life sciences. Our company is organized around the concept of the Veradigm Network, which provides life science companies and health plans access to de-identified patient data, provider connectivity, and analytics at scale. Through this network, we have the unique ability to connect data to analytics to enable high quality, lower, lower cost care. Now, our core assets center on Veradigm's network of over 400,000 providers, with de-identified data spanning over 200 million patients.
Our data set is large, but it is larger than just us. By securing access to proprietary registries and expanding our partnerships, we can continue to service our customers, but more importantly, generate large-scale, high-quality data. And we do not take our position in the market lightly. Rather, we understand the responsibility to be good stewards of healthcare data, which comes with our size, longevity, and expertise. The high quality, large scale, rich clinical data at our fingertips can be monetized and analyzed to unlock key insights in real time. These insights can advance care options, decisions, and drive value-based care. And we know that data is the lifeblood of clinical research, and we have the ability to improve research, transform healthcare, and provide data-driven solutions to our customers. Now, this marketplace has actually been fueling our growth.
The demand for better data continues to grow, we all know that, because we rely on data to determine care, cost effectiveness, new methods, and research. And Veradigm has been able to meet this demand for our customers with our scale, connectivity, and expertise, and we are well positioned to continue to grow our core business of supporting physicians and providers, payers, and life science research enterprises. But there is more. What I have described in the few slides before is only one level of value that we are at the center of, because we are actually at the center of a much greater transformation in healthcare... and we have the opportunity to support both our customers and the entire healthcare ecosystem in new ways.
In the last 20+ years, our work has supporting physicians, practices, providers, and even our work in the past several years, supporting payers and more recently, life sciences. We have created a unique and valuable data set that positions us to be the health intelligence leader for the future. Given the rapidly advancing state of technology today, the moment is right for us to be thinking about how we can leverage this high value, real-time, clinical data sets and clean records, which we have to shape the future of healthcare data intelligence. With the proper use and analysis of this data, including social determinants of health , we can provide highly differentiated and more advanced solutions to our customers. Why are we so strategically positioned to lead this next generation of health intelligence products?
Well, we are present across the ecosystem of healthcare stakeholders, from payers, providers, and life sciences. We have incredible high quality, high value, large-scale data, driven by over, again, 400,000 providers and 200 million patients. We are already delivering analytical value today. We have the accumulated expertise to supercharge development of a next generation of suite of products capitalizing on AI technology capabilities. In fact, with the advent of new tools, technology, and generative AI, we are positioned to be at the center of the largest platform shift in healthcare, and this is a shift that I remember very clearly that was predicted over 20 years ago when I first stepped into this space and industry.
As I mentioned earlier, I've spent my career in health information technology, and I have seen the evolving shifts of focus from making electronic medical records in the late nineties to today, as we try to answer big healthcare questions across disparate sources with technology and advanced analytics. But underlying all of the advances in our industry for the past 25 years was a buildup of data, even though it's been of varying quality. But today's tools can unlock, clean, and bridge many of the gaps in care and research, providing us with huge opportunities to learn and provide better care. And I am so proud to say that Veradigm is the only independent EHR and EMR provider with clean and complete data that spans clinical and claims.
With our stewardship and vision, we also have the ability to understand research gaps in real time, solve them, and then build solutions that help our customers capture further value. Now, you may be wondering, is this a new strategy? And the answer is no. This is not a new strategy. It is simply a step in the larger vision of the company, which actually began back in 2022, so almost two years ago, when the organization focused its energy to open up two new markets and start this journey. This is a continuum story, and we have the unique opportunity to be the health intelligence provider for the next century of care. Sticky pages. Sorry. Veradigm is more than a software provider to physicians, practices, providers, payers.
Its future lies in how it can capitalize on being a data asset with great data stewardship of the past and present, to support all of its customers as it expands its role to support the narrowing of gaps of bench to bedside, bedside to home, care to research, empowerment of patients and physicians, and care and health equity. And we do this through defining as well as providing intelligence. And if we do this well, in fact, I know we will do this well, we will provide valuable data products for our current customers and future customers, and we will drive improved and lower cost healthcare outcomes. We have the opportunity to lead the way with a fundamental understanding of physicians, physician workflow, providing care, supporting accurate payments, and providing value to life science research.
And we've been investing in technology, people, workflow, process all along, and we have been learning and discussing the benefits and perils, and yes, I say the perils, of new technology such as generative AI and the ethical and appropriate application to healthcare data. This shift is happening, and Veradigm is at the forefront. We have the right team, the right data, the right technology, and all at the right time, and we really look forward to sharing more as we build this future together. Thank you. Actually, and that ends the presentation aspect. We're going to now have a session with Anne and Will, sorry, and with Will Manidis, the CEO of ScienceIO. And given the exciting position that we are in capitalizing on AI, we actually thought it was really important to bring in one of our collaborators to talk further about this opportunity. So thank you.
Great.
... Welcome, Will.
Thank you for having me.
So, you know, maybe we could kick it off here with a discussion on some of the risks and perils of, you know, general purpose models in healthcare. So, you know, we've heard a lot about large language models. You know, are these a good fit in healthcare?
Sure. It's, it's undeniable that ChatGPT was kind of a break glass moment for industry, right? We went from having 20 years of AI promise to having a technology that is in everyone's hands, that is intelligent and broadly applicable. The issue is we haven't benchmarked that technology or built it in a way that is responsible for healthcare. The models that we serve to customers are a reflection of the data that they're trained on, and as we move from prototypes to production to patients, it's important that we reflect on the data these models are trained on, how we deploy them, and the safeguards we build.
At ScienceIO, we focus on building the world's safest large language models, like we used in healthcare, and our collaboration with Veradigm is focused around getting access to high-quality data across disparate patient populations to ensure these models can be used safely and well across industry.
Absolutely. I mean, I think one of the things that people often talk about is large language models, but what they aren't talking about are small language models, which is this idea that the healthcare data is incredibly specific, and it has its own strange and wonky ways of being presented. And so having the ability to understand what that means and recognize that every single data point, every observational data point is connected to a patient, is incredibly important and creates an extreme burden in some sense, and in responsibility to make sure we do this right. So, we are very excited about this collaboration, too.
Yeah. I would just add that we often forget, working in healthcare, that healthcare is a different language, right? The jargon we use, the terminology, all of the stuff that seems easy to us is entirely foreign to the training sets of these models. So as we think about building specialized healthcare intelligence, it will require deviating from these general purpose models to models that have been trained, benchmarked, and built for this use case.
Can we discuss maybe some of the pitfalls of, you know, using these as it relates to healthcare space? I think, you know, something that we've heard from others as they think about building out AI models in healthcare is, you know, hallucinations. You know, that, you know, things can appear, you know, differently than they are, or, you know, there's a real problem with health equity, you know, and making sure that you're getting information across a really wide spectrum.
Sure. I would say as we move into production across industry, many folks are deploying generalist models and seeing very strong results because they're piloting on small patient populations that are largely homogeneous. As we move into healthcare broadly, and we see patients that are as diverse as the folks in this room and more, having these models be safely deployed across that context is incredibly difficult, and being able to benchmark where the shortfalls of these models are is largely a problem that I don't think the industry has taken seriously enough, right? Hallucination is step one, but step two is the bias in the underlying data, the bias in the kind of observations we make, and the patients we see, and the results these models put out.
I think that's a really important point because that's the one wonderful thing about Veradigm, is if you think about where our physician practices are and where our 200 million patients are, this is actually a really well represented of all the diversity, both socioeconomic as well as from a racial and every other way of thinking about diversity. We have this patient population, and we have the observed data, which is incredibly important in a time when you're building models, because anytime there's a hallucination in the mix, okay, and I'll use that term, and that way is that you suddenly have not only a data point that's no longer an observed data point, it's now attached to a patient.
Which means what happens if you start thinking about what will occur when that piece of data starts to follow a patient, and what will happen if you go the other direction, where it gets included in analysis that help generate research results that would affect many patients? And then how does the responsibility, auditability, liability, all work around something like that? So you can understand that in some sense, being very careful and being very ethical is incredibly important and is a, you know, as a steward of all of this data, this is why we have been so excited about taking this forward in the right way.
Maybe, you know what? Another one is, you know, how are maybe both of your companies, you know, kind of positioned for, you know, the future of AI in healthcare, and also, you know, how are you prepared to, you know, maybe close the gaps in healthcare for some of these underserved populations?
I think part of it is really it comes down to analysis. Most of the research that we do today is still disproportionately based on groups of people who don't necessarily represent the diversity of how disease plays out. We are still trying to understand how different subpopulations are affected by different therapeutics and by different modalities of care. And because we don't have a good understanding of it, it becomes all the more important that whatever models we build are gonna be built on data that are come from these different individual subpopulations, so that we can start to think about what are possible better treatments and also track better outcomes.
Yeah, I would say every judgment we make in healthcare analytics is a reflection of the patients that come into the clinics that we observe, right? Large AI efforts in healthcare to date have been built around centers of excellence, whose data skews meaningfully from the kind of populations that Veradigm sees. Right, by opening the remit of possible data that these models can see, train on, and understand, we can make more accurate predictions across the long tail than purely representing a very narrow segment of patients that have been served today. As we see that, it's also important to have the workflow conductivity to be able to push those tools into the tools that providers are already using.
I think it's not news to anyone here that providers are not in love with uptaking new tools, so the ability to give this tooling and infrastructure in the places where the care is already being delivered, rather than moving them into new ecosystems that they may not understand in clinics that are already overstressed and overpromised, is incredibly important.
Maybe we could spend a little bit of time, you know, addressing some of the revenue models for applied healthcare AI.
Sure. I mean, one of the interesting things about revenue models is if you provide better care, there is already a revenue model there, because you are already starting to think about not only the information and how it may save in making better decisions, but you are also collecting information that becomes useful for understanding treatments in the future and as well as treatments at that moment in time. So I think that what's nice about it is to kind of think of almost every treatment and every healthcare choice actually is a generation of data, and that data, that observed data, becomes useful not only in the application of caring, of which there are many revenue models about better care, or it is also...
Not really an or, it provides actual support for better and better research, tighter research, more narrow, more specific research, broader research, and all of these do come with their own revenue models that are oftentimes driven around the type of outcomes that come from it or the type of analysis that comes from it.
Yeah, I would just emphasize a really exciting part of this is the refocus on patients as the unit of value. We have been focused over the last decade a lot on system-level change and not enough on the uniqueness that every patient brings when they enter the clinic. And by training generative models that reflect the underlying characteristics of patients we see, the most underserved, the most rare patients, are the most valuable to the healthcare infrastructure providers. So by being able to see such a wide remit of patients, we have a much wider and deeper incentive to serve the patients along those distributions of the long tails, rather than focusing on the core demographics that we see in centers of excellence.
You talked a little bit about, you know, before, about, you know, kind of providers being slow-moving and sometimes resistant to change. Can you maybe spend some time talking about, you know, what are the biggest hurdles to creating some of these, you know, intelligent healthcare tools?
Yeah. I don't think it's provider excitement. Honestly, if you go on TikTok today and type in ChatGPT doctor, you will see physicians using ChatGPT for their daily practice, from submitting prior authorizations to filling out documents. Providers want this tooling. The issue is getting that tooling in their hands in a way that is safe and reasonable for their actual use. I think a really funny example is if you search that on TikTok, the most prominent video is a provider filling out a claim denial form, and the citations that ChatGPT references are actually entirely hallucinated. They're not real insurance policies.
It's not that providers are skeptical this tooling or difficult to take it up, but it's they're going outside of the tooling they already use and often doing so in unsafe ways, rather than being able to serve them responsible tooling in the infrastructure they already use.
I think history hasn't been on our side, right? Most of the tools that we have given physicians over the years have not been developed by physicians or by the people who are in healthcare. They've been developed by technology companies who may have treated a lot of the different types of applications very much like a vertical. And while that has worked for the most part in a first level and a second level, now we're going down to the point of a patient, and now we're going at a point where it took us 20 years to get to this point. It almost is that no real technology company by itself should be able to do this, but an organization that is a healthcare tech company can go in the right direction.
Maybe just, you know, one more on AI. You know, how do you envision, you know, the future impact of AI on, on healthcare?
I think a lot of how people have been thinking about it to date is giving kind of additional leverage to the workflows providers already do. I think that is both correct and also an understatement of the total impact that it will have. The kind of intelligence we're building with large language models are very different from the kind of intelligence that a provider might have and has different strengths and weaknesses. If we think about the kind of population scale analytics that were previously impossible because the data wasn't ready or well formatted enough, that's something a provider could never do. No provider is gonna sit there and abstract thousands of medical records by hand, but a model can do that easily.
I think about both increased efficiency, safety, and cost at point of clinic, but also opening up new analytic pathways and new modalities of care that are possible by having algorithms that can do things that are not things providers can do.
And we can give them back those tools. We can give them back the data. We've asked a lot of data from them over the years, but we haven't necessarily given back them the type of tools to use their own data to help support their patients, but also support the research enterprise in general.
I mean, it's funny to think, I mean, you know, the EMR is such a source of friction and burden for the provider, you know, in the healthcare ecosystem. It's kind of an interesting perspective to maybe, you know, fix that problem.
That's right. Give it back.
You know, maybe just, just one more for you, Dr. Ho, on just, you know, kind of Veradigm specifically. What are you most excited for as you look into 2024, now sitting in the seat where you sit?
I'm super excited about 2024. We have been waiting for this moment in such a long time, and I speak about that more historically and also personally, too. This is a moment where we have an understanding of what data is... Quality of data. We can actually discern the different levels of quality of data, and we also have an opportunity to model at a speed we've never modeled at before. We actually could propagate and augment people's capabilities in a way that hasn't been dreamt of before, and we can do it in a really deliberate, very thoughtful, very ethical way. I'm super excited about what we can do. We're in a great position.
Terrific. Well, thank you so much for joining us today, and, and thanks, everyone, for attending.