CareCloud, Inc. (CCLD) Earnings Call Transcript & Summary

Okay, guys, I think we're going to go ahead and officially begin. So let me do my thing. Hi, everyone. Thank you so much for joining us. We appreciate you taking the time out this afternoon. Before we kick this off, I want to point out that we are going to take audience questions. So please utilize the Q&A section at the bottom of your screen. Type in questions as they come up, and we'll hold those to the end and go through them. I would now like to turn this over to Harvey [ Briggs ], our Chief Communications Officer here at Forest Family Office, who will be moderating today's panel. So thank you, everyone.

Great. Thank you, Kelly, and thank you, everyone, for taking the time out of your busy schedule to be with us today. There are a few categories hotter in tech right now than AI. Yet it may surprise people to know that AI has been used in industries, including health care for the past decade. One company that has pioneered the use of AI in health care is CareCloud. It's a technology and services company that helps doctors, clinics and health systems focus on patients by providing them with the tools and support they need to run all aspects of their business. Joining us today are 2 gentlemen who know a thing or 2 about this topic. They are Bill Sutherland, the Director of Research at the Benchmark Company; and Bill Korn, Chief Strategy Officer for CareCloud. Let me tell you a little bit about them. Bill Sutherland is an equity research analyst focused on health care services and digital health. His coverage focus includes companies providing virtual care and remote patient monitoring, value-based care and post acute and alternate site health care services. He was previously a senior research analyst with Northland Securities, Director of Research with Boenning & Scattergood and a senior research analyst with Janney Montgomery Scott. Bill holds a BA from the University of North Carolina at Chapel Hill and an MBA from Columbia University. Bill Korn joined CareCloud in 2013, serving as Chief Financial Officer until his appointment as Chief Strategy Officer just last month. Bill has served as CFO for 6 other businesses, including Snap One and Antenna Software, where he enabled the business to grow at an 87% annual rate through a combination of acquisitions and organic growth. He also spent 10 years with IBM as a member of the senior management team that created IBM's highly successful services strategy. Bill graduated [indiscernible] from Harvard and received his MBA from Harvard Business School. And with that, I'm going to get out of the way and turn this over to the experts, Bill Sutherland, please take it away.

Thanks, Harvey. I never kind of think about where I went to school. But it's good to be reminded. So generative AI, right, it's like such a catch phrase now and I think we all know kind of what it does. It's been doing it with numbers and data sets for a long, long time. It's begun to do it through this large language module technology through to handle human interaction, follow-up questions and can actually challenge premises that seem to be incorrect and reject inappropriate responses, it's not -- it's early days. It's still a little bit of a toddler. It assumes it knows things [but it] doesn't, but it gets smarter every day, which is -- what's amazing about it was a little anxiety [provoking] about it. Just to give you an example of how much it's gotten or how far it's gotten in health care, it recently passed the U.S. medical licensing exam. And it's actually prompted some of the major scientific journals to ban or restrict its use in its research. So already some cautionary tails out there, but also some amazing use cases. Investments are accelerating too in the technology in 2022, prior to the release of ChatGPT, healthcare AI investments totaled $4.4 billion according to Rock Health, and that's probably going to be double that this year. with the new initiatives going on. So there seems to be a new use case almost every day. As I've kind of looked through the various journals and so forth and just the headlines. I think there's 4 broad areas that you can think about where ChatGPT -- well, generative AI is going to have a pretty big impact in health care. I think first and foremost, at least first, maybe not foremost, but seeing a lot of impact in areas involving patient and clinical documentation, improving the workflow basically, improve clinical -- clinician productivity, better manage really what health care is dealing with today more than anything, which is a staffing crisis in some cases, at least staffing shortages in most cases, and with also a focus and that's on that score with reducing clinician burnout, which in turn is adding to the staffing shortage problems. So in this area, there's 3 applications in particular. One, I think that's gotten a lot of attention is sometimes called AI medical scribe, which is, as it sounds, replacing the scribe [indiscernible] implementing a technology, the supplants the human scribe, that's sometimes used in cases where the note accuracy is really important. And Here, you have a couple of companies, the biggest that you probably have heard of is a Nuance, now part of Microsoft and a couple of small ones on public called [Augmetix]. And they are -- they've -- Nuance has always had a dictation capability, but the technology is now moving towards capturing ambient conversation. And then structuring notes from that conversation and structuring the notes in such a way that they can be accepted into EHR, the electronic health records. So [Augmetix] has documented time savings of up to 3 hours for the physicians that utilize -- 3 hours a day for the physicians that have utilized their system. The next big category, I think, is patient communication, often inside the health system portals, where personalized messages can be sent to the patients without [that] human action. These are messages as simple as upcoming appointments, provide intake forms but sometimes getting into -- answers to pretty basic queries from the patients that they might put into the message part of the portal. That would be saying I've got an elevated heart rate or I need an antibiotic prescription. And so automatic -- automated responses are -- have been created tools have been created for that, like Epic's AI-based My Chart. And then there's just lots of small benefits in this patient and clinical documentation area. And I'd like to give an example, just to show you kind of where this can have a meaningful impact, although it sounds like it's a very small application. Baptist Health in Florida has developed an enterprise AI chatbot for password resets, and basically, the calls inevitably come in because people forget their password, they've reduced the enquiries from taking around 9 minutes to a few seconds. So you think about just the magnitude of that time savings. So that's the first use case. The second 1 seems to be in actual treatment, more accurate detection, better disease management, the primary use for now is in precision medicine, and that's detecting abnormalities like [health] scans that might be missed by the radiologist human eye. It involves robotic-assisted surgery, it involves the clinician assistant to a physician where that assistant is now digital and driven by AI. There's one example, I've been reading about a lot called Tempus One that is a voice and text as system and can actually access new clinical test reports on-demand, receive status updates, filter the patient data according to genetic alterations and their tumors, summarize patient information and provide clinical guidelines. And this is all just driven by a Q&A. It can either be verbal or taxed by the physician. It's a huge time savings and they're finding that in some cases, they're getting more accurate diagnosis. The other big category, number 3 would be drug development. Large language modules here has the potential to speed up drug development by cutting down the time it takes to review the scientific literature, identify targets, identify clinical trial volunteers. So I don't have any stats there, but I think that is going to be a major focus of the application or time because that's a place where you could see just a huge gain in productivity and shortening those trials, which it's all about getting to market. And then finally, the fourth 1 would be in the EHR and revenue cycle space. And this is really CareCloud's core focus to date. And so I think we'll shift over now and talk a little bit about CareCloud, I'm going to ask Bill Korn to [know] the products so the people have an understanding of what we guys do and then we'll move on to the AI part of it.

Thanks, Bill, and thanks for a good introduction. And as you say, everybody is talking about AI today and hopefully, by the end of the hour, people will feel like they have a little better facility with what it is and what it can do for them. CareCloud is a health care technology company. I joined about 10 years ago, we went public 9 years ago. We were a $10 million company at the time of the IPO. We did $140 million of revenue last year. And our business is helping doctors primarily practices, small practices, large practices. We also help hospitals and our goal is really to do the things for them that allow them to focus on the patient. So making sure that the claims are submitted and paid, giving them the EHR tools to take notes, they practice management systems to keep track of appointments. We added telehealth to our product suite back in 2019. Nobody cared about it in 2019, but I'll tell you in a couple of months later when COVID hit, I will say that our customers were really pleased to know that -- there was a HIPAA-compliant telehealth app that was built into the platform, right into their EHR, right into the patient portal and they could continue to do the practice, and we can continue to get them paid. So I think we've been somewhat on the leading edge of using technology for a variety of things over the last 22 years since our Carecloud's has been in business.

So when did AI become part of your product or your tech stack, as I say?

Yes. So we first started using AI back in 2005. So people think about that being a little bit on the early side, although I'll date myself in the 1990s when I was at IBM, we were thinking about AI, but never really getting it into full use other than trying to play chess. But CareCloud back in 2005, focused on what's called claims scrubbing. So 1 of the things that's important for doctors is to be sure the claims get paid and how do you do that? The best way is to submit the claim with everything that the insurance company needs. Generation 1 was people pouring through the records and figuring out, here's what Blue Cross wants in New York to do this kind of claim. But that becomes difficult to staff up and do that. So back in 2005, we started using AI to look for patterns and to figure out what did you need and to be able to pull the data that was necessary so that when you submit things, you maximize the chances that it could be paid the first time. And today, I'm pleased to report that we have something like a 99% first pass rate, meaning 99% of the claims get paid with no human interaction. And that's great, and it probably would be impossible to do that well with just people, and of course, would be much more expensive. So thinking about using AI, I mean, we've been using it in our business. People don't necessarily know about it. They don't think about it, all they really care about is I'm the doctor and when I'm using CareCloud, my claims get paid and they get paid quickly. And how it happens, it doesn't really matter.

Is that the primary where -- is that the primary place that you're seeing it applied?

Well, that's -- I would say that was the first place. I mean, that was the first place back, I don't know, 18 years ago. I mean today, we have 50, 75 robotic process automation bots that do all kinds of different things. And again, another example is checking insurance eligibility. So -- if you think back generation 1 is the person handed a Blue Cross card to the practice when they checked in, but maybe that was when they checked in last month or the month before. At the moment that they're actually seeing this visit, you got to go check it. And so again, without using AI, you're trying to run the card and it doesn't go through and you're like okay, the person isn't insured. Well, now we use AI and you try to say, okay, this card was -- didn't go through, what could that be? Could the person have switched insurance companies? And you try to sort of think about where could it be where else could I find it, "Oh, I found Bill Korn. He wasn't with UnitedHealthcare, he was with [indiscernible] -- he was -- and he tried to use the system. So we're using AI to do a bunch of things. Most of them are not -- are shattering things. They're not going to change people's lives. But again, anything that you can do that can automate the process, remove some of the costs and remove some of the potential mistakes. That's got to be good for everybody.

So are you -- as far as product decline in the EHR and revenue cycle space, it's mostly your ability to improve the productivity or efficiency of the revenue cycle because that's a managed service that you all do as opposed to the software module like in the [CareCloud]?

Sure. It's a combination. I think we started with automating the things that our team is doing for the practice, which includes all kinds of documentation, it includes answering questions. I mean insurance asked the question or the practice asked a question or maybe even the patient is asking a question about their bill. If I can use AI to do that, that's great. We've got new services that are starting to take advantage of it. So for example, we introduced last year something called remote patient monitoring. So this takes advantage of rules that have changed with Medicare and private insurance recently. We [have] people with chronic conditions can have visits, remote visits, and they could also get devices that are paid for by insurance that will check things for example, like your blood pressure. And again, what would you do if you were at doctor's practice and 500 of your patients called in their blood pressure readings every day. I mean there's nothing you could do with that. By the time you figured out how to handle it, the day would be over. On the other hand, I could put the -- i could put AI into this. And not only are we putting this into [your] chart, but it's looking at your pattern. And it says, okay, this is Bill Sutherland's number today. That's perfectly normal. Everything is great. This is Harvey's number. Well, maybe he needs to actually go and talk to the doctor, maybe he needs a change in Medication. So again, I'd say a service offering like that without AI would probably be impossible because otherwise, you just generate gobs of data that wouldn't be actionable. And when we think about kind of where we're going to go here is for AI to really help the doctors to improve the patient experience. And our approach is going to be helping the doctors -- we're not going to displace the doctor. We don't think it makes -- we're not yet at the stage where people are going to go to [indiscernible] or [ R2D2 ] to get their medical care. That's not where we're going. But again, a process that doctors do all the time is something call the differential diagnosis, they teach this in med school. You take the history, you look at the symptoms, you come up with some potential diagnosis and then you come up with how am I going to figure out which 1 is it? What questions do I ask? What tests do I run and then I could figure out what the right thing this. Well, Okay. The doctor learns that methodology. But to do it right, they need to kind of think about all the potential things and it's error prone. But now you say, "I could use AI. I could have a virtual assistant. And again, we're not going to rely on that virtual assistant to do everything, but they can prompt the doctor, here's what you should ask about. And the doctor may say, you don't need to do that. I know in this case that, that's not where I need to go. It's better to have brought to their attention than for the -- to hope that the point that they're seeing that patient -- they aren't distracted. They aren't confusing 1 potential disease for another, they aren't looking at the most obvious things. So again, we look at this and say, we're not ready to have people delegate their health care to AI. That's not prudent. But to the extent that we can help the doctors and their staffs, I mean, they're all having trouble getting enough staff. So -- and every doctor needs more experienced nurses. Okay. So now I have a virtual nurse that can help your regular nurse be a little more productive. Maybe that regular nurse isn't burned out, maybe they feel like they'll stay a little longer. It's got to be good for everybody involved.

Do you have a clinical assistant product that you're commercially putting out there?

We don't have anything that's out there now. And again, I think that 1 of the things that people have to have a little bit of patience. The day that you read about it, the day that you see it on YouTube, that is necessarily the date that it's fully there and ready to go. And again, we've got 18 years of experience using AI. It wasn't always easy. And I think people's attention is really drawn by generative AI because finally, this is like the moment in 1996 where Motorola introduced the StarTAC cell phone and people like, wow, now I have the same communicator is Captain Kirk. And they get the same perspective, like, wow, the computer can now look at things in analog, well, you know what, even on Star Trek, the computer wasn't right all the time. And at the end of the day, Kirk needed to go talk to the people that are working for him to figure out what to do -- and -- so you got to think about it. It's not going to go from zero to it's doing everything. But on the other hand, you can do a lot of things today and it's going to get better at doing them over time. So challenging giving it more opportunities is only going to be helpful.

You all thought about the risks that occur by deploying AI for the end user more than anything, I guess?

Absolutely. And I'd say Step 1 is protecting your data. And it's great to think about using a free version of ChatGPT, but doing something like that is going to take all your data and it's being exported to the world. So that clearly doesn't make any sense. And it also means that it's been trained with data that you can't control, which means some of that data may not be real. So the first thing that's clear in any health care application is you need a private session. You need to be paying for a tool with its own container, with data that's yours, that stays private and Google and others do this and they've set it up in a way that it's mathematically impossible for them to get in and see what you've put much less than anybody else. And that sort of Step 1 is if you don't do that, you shouldn't even be thinking about AI. Step 2 is be thinking about what's the problem you're trying to solve. And how do I use the technology and use the people's expertise together to get me the best solution to the problem. And -- and that sounds obvious, but again, I'll use an analogy. I sometimes have seen people when they're putting together a business PowerPoint presentation. They think they have the greatest slides, and they're done and they have 20 gorgeous slides but that doesn't tell the story, and it doesn't answer the questions that the decision maker needs to decide what to do. So you're much better off saying, what do I need if I'm going to get out of this meeting, and I'm going to get the answer I want, what are the 2 or 3 points I need to make? And let me put together a couple of [pointers] like to address that, and they don't have to be pretty -- they just have to be self-explanatory. And I think the same [issue] with the technology, asking everybody -- because that's not going to get you anywhere. But based on your situation, what are the -- what's the right line of questioning, what's the -- what are the possible treatment options, then you could get to something that makes some sense. And of course, the last step, of course, is when you get the answer, you need to evaluate it. And again, I'll use the analogy of the navigation system in my car is good, some of the time, most of the time. But what it tells me that the Chinese restaurant that I know is 2 miles away, is 5,000 miles away in the South Atlantic, it's wrong, it's completely wrong. And I'm not going to follow the car. I'm going to go back and use my common sense. And again, you need to do the same thing with generative AI. If you're getting an answer that just seems like it's wrong, then it's wrong and don't go and don't go and just proceed blindly on it.

So was your -- with your CareCloud product set that came with the acquisition that's primarily HR. Are you all adding functionality on a new version. Is that kind of your approach to rolling it out on that side?

And we're not ready to announce new product versions, but at the point that we bought the company whose name we took back in 2020, CareCloud was working on a new version and we've continued that development effort. And we will, at some point, have a new version. I mean we've got a couple of well-acclaimed certified EHRs already, but there is another version that's in the works that's going to have more. So I guess I'd just say stay tuned.

Okay.

The best and brightest are working on it, and it's not me.

So we talk about the what if's -- how do you think about the limits of the AI technology. I mean -- it's not going to change everything we do, right?

It's not going to change everything we do. And I think again, like with all technology, you have to think about it, and there's this garbage in, garbage out principle. I mean if you -- if it isn't starting with good information, no matter what you do, you're never going to come up with a good result. So you got to be thinking about what is it that you're giving it and you've got to be applying good common sense. And if the result is counterintuitive, then you're not there. And I think people will often -- you've seen newspaper articles about people asking an AI silly questions where you'd say they knew that the AI wasn't going to give a [rather] valid answer. And they asked it anyway. And of course, they got a ridiculous mumbo jumbo answer.

Garbage in, garbage out.

You got to think about what are you asking it for? And if you ask, it's something that's within the realm of the data that it's been trained on, then it's got a chance that it could do something for you. And it's never going to predict what number is going to get chosen in the lottery tomorrow. So you can ask it that question, but it seems to me you're never going to get the right answer. So it's not even worth trying.

And a lot of folks worry about this is going to be automation running out of control that eliminates not we've had jobs get eliminated through the industrial revolution and robotics on the assembly line and in those kinds of tasks. But here, we've got something that could be replacing a lot of professional kind of working roles. How do you guys at CareCloud see it?

I mean as you said, jobs have been automated over the years. I mean think about if you were building the pyramids again, I mean think about those heavy blocks, how you would have gotten them up to the top, you wouldn't have done it the way they did it back in those days. And you think about how many people were building Stagecoaches 200 years ago and how many are you doing that today? I mean -- and I mean, even a mundane task. I mean everything that can get done with brute force can get done better through automation. And that's true if the brute force is manual labor, but it's also true if it's even sort of simple cerebral kind of things. Again, I mean how many people go through the act of stock picking today starting by opening up a printed newspaper and starting to go through print -- I mean, no, you're going to go into a database, you got to set up a screen, you're going to look at things. You're still going to have to -- at the end of the day, you're not going to put $1 million on whatever your sales person tells you to do, but you're not going to start from ground zero. And I guess my feeling is if we don't deploy AI, somebody else will and the people in our organization who might -- their job might get automated, okay, that's going to happen whether we do it or not. And it seems to me -- the right way to think about this is how do I make the people more productive because we're growing, and we intend to continue growing -- so the goal would be how do I handle more revenue and do more work for more customers with the team I've got. And telling the existing people, you need to work more hours, that's not going to work, telling them you need to do it faster, that's not going to work. I mean so -- and we're telling the customers, we're going to just cut the corners, that's not going to work. So it seems to me, how do you let the customers get more done for less and how do you let the employee handle more work without getting stressed, it's by making people productive and using the technology. And I think that what's our job as management, it's to make sure that we help our people get as productive as possible, and help our investors get a better return on investment because we're a more profitable company. And using AI is one, certainly not the only, but it's certainly one means to help us there. Doesn't hurt the environment. So there's a lot going for it.

Well, like [indiscernible] to which you've got another server [farm] out there.

But it'd be worse if I needed to hire another 10,000 people and pay for the housing and the food, so to the extent that I can automate it, I can do it more cost effectively and better for the world.

Don't you think it's rapidly becoming table stakes that you're going to be -- your competitive profile is going to deteriorate so quickly, I mean that's how people are going to lose their jobs if CareCloud [indiscernible] keep up?

I think that's certainly true. If you're a company and you say, "I'm going to" -- you're going to resist this and I'm going to keep doing this stuff, and I'm going to do everything the way I did in the 1960s. Exactly.

What specifically with so much of your workforce offshore, do you think your margin profile will be impacted by the shift?

Well, I think again, if we weren't thinking as we always are, if we weren't thinking about how do I continue to be more productive than the margins would absolutely be impacted. And I think that if you can do more with less, whether that's through automation or through a better offshore team or whatever other means that allows you to give more to your customers every year, and you can decide how much do I -- the benefit do I keep? How much of it gets passed on to my customer. Sometimes it's actually benefits to passing that on to the customer and that improves the customer retention and satisfaction and helps you get new customers. Sometimes you say, okay, it makes sense to keep some of this myself. But clearly, anything that you can do that can improve your overall efficiency. At the end of the day, it's good for the bottom line.

Well, you've successfully launched now both [real] patient monitoring and chronic care management. Why don't we [indiscerbile] initiatives that are highly dependent on the delivery of a service being AI driven. I think it'd be interesting to hear more detail since those -- those are going to be as far as your outlook, key pieces of the growth acceleration that I think you guys are seeing. So talk a little bit about the rollout of those 2 and maybe the margin profile that you're potentially going to see and how AI kind of [indiscernible] into that?

Yes. Yes. So we launched chronic care management and remote patient monitoring last year, and we've gotten great reception from the doctors who were our customers. And today, we've tended to focus on deploying to our existing customer base, eventually our thinking is that we'll offer these services to others, we've kind of designed the services in a way that they can play nicely with other EHRs so they can integrate with ours, but they can integrate with the third party. And we're not going to tell doctors, you have to switch to us in order to be able to use those capabilities. I'd say unlike other service offerings that we've had in both the chronic care management and the remote patient monitoring situations, you need that not only the doctor to say, yes, I'm signing up with you, but you need the patients as well. And so what we're seeing on a monthly basis is a dramatic increase. I don't know if it's doubling, but it's pretty close to that on a month-by-month basis of how many visits, how much revenue is it deploying again starting from zero. So when you start from zero, you don't start to see a lot, but we're on a path where we are expecting to see some hefty revenue contribution in the second half of the year. And I think the rollout is starting with chronic care management, which is really a little bit less technology because as a person involved in the chronic care management. You've got a care manager who is connecting with the patient on a monthly basis remotely, could be through video, it could even be through audio and you're dealing with patients with chronic conditions, making sure that everything is okay. And the alternative, again, if you had diabetes, you're not going to go see your endocrinologist every month. You're. Going to think you're fine until 1 day you're not, and then you're going to wind up in the ER. So if we can check you on a regular basis and catch when there's when there's a challenge when the numbers don't look right, when it looks like you're not being compliant when it looks like we need to change the medication, the dosage, let's get you before you're in trouble before you even know you're in trouble. And yes, then we can ward off the $10,000 ER visit and the weak spent in the hospital that you really weren't looking forward to. So -- so we're excited to see that rolling out, and you'll hear more about that during the second half of the year as we continue to progress.

Those are obviously growing faster than your other core businesses. And so will the overall margin be beneficial as it gets bigger?

Yes. I mean, I think that the -- I think the overall margin potential is probably bigger in the remote patient monitoring, which really utilizes the AI more so than the chronic care management where I'm going to need to be paying people, whether they're my employees or even subcontractors. I am going to need people on the phone, and we've chosen -- or on the computer, we've chosen to set this up in a way where we're trying to encourage the use -- so we're not trying to be greedy on the margins there. But I do think that when I think about remote patient monitoring, my cost is the cost of the technology. And so once I'm getting to scale there, I think that will certainly help overall margins.

That could be closer to your software side than is what you're saying?

People often try to figure out how much margins the software versus the service and to us, in some ways, this is like how much of the cost of the last Mercedes that you bought, how much was the engine and how much was the body? I mean at the end of the day, you're providing a service for the doctors. There's a software, there's a people [indiscernible], but it doesn't really matter.

It's sense. Got it.

You bundle it together, you help them to be better in their practice of medicine. And that's the name of the game.

Should we see if there are questions?

Absolutely.

Yes. Yes, it looks like there are several that have come in. So I'm just going to go from the beginning that -- so someone had commented that it's unreal that your market cap is still only $50 million, especially with the rush into AI, what's the disconnect?

So I guess the first thing is when you -- I mean, anyway, like every other public company, no public company ever thinks this market cap is appropriate. But -- but I think the right measure for us, you got to look at enterprise value because when you say $50 million, that's the market cap of the common, there's another $150 million of preferred -- so I think the right number is to look at the overall enterprise value of $200 million. And I'd say this is a good example of ChatGPT and others there are sources on the Internet that say, "Oh, your market cap is $50 million, and they're wrong. If you look at Bill's report or any of the other analysts you follow us, they'll all look and they'll say, this is -- when you look at the total value of the company, all 3 classes of shares and subtract out the value of the cash, there's a $200 million enterprise value, which, again, Bill and the other analysts who look at us, we'll still say, gee, $200 million, that's still a small number. It's still a lower multiple of revenue, a lower multiple of EBITDA than your peers. But I think you have to look at the 200 number, not the 50 number. And if you're just looking at the 50, it makes the case seem more compelling, but it's not realistic. And so if we could figure out a way to get a feedback loop to get those numbers correct so that when people looked at Yahoo and other places, they actually got real numbers, that would be great.

Can you touch upon your competition? And what are your advantages?

Yes. So I mean, competition, you could look at others who are in the health care technology space. You could -- whether it's the Epics and the [indiscernible] of the world, who are more in the hospital space, the [Allscripts] and th NextGen or you could look at the technology companies or the Googles of the world and say, okay, at what level, they're all potentially providing some of these capabilities. I think what we have that they don't have is we've got this stable base of customers and a lot of great data that relates to that, that needs to be well protected. But there's a lot that, again, not through a public generative AI, but through creating carefully curated data sets, a lot of things that could be learned from that. And in fact, our CEO, Hadi Chaudhry, spent some time with Google's Head of their AI business over the last couple of weeks. And along with the CEOs of other companies who are potentially using AI. And I think it became clear in those conversations that we're actually ahead of a lot of the other people there in terms of thinking about this and have a lot of capabilities that we could leverage. And again, I think this is ours to continue to take advantage of.

So I think you just touched on this, but how does your approach to AI differentiate you from other companies offering clinic management systems?

I mean if you're just offering a clinic management system, and you don't have 40,000 doctors who are utilizing some of your capabilities every day and you don't have information, you don't have health regular -- what are you drawing information from. It's the data that's out on the Internet. I mean, that's not going to be a very good database to train the AI to actually help the doctors. On the other hand, if I can start with -- all right, let's imagine that I'm helping people who are dermatologists. If I can start with data of here's what's in records of other patients of dermatologists and here's what symptoms look like and here's what patterns look like. And here's what people were diagnosed with, okay, so now what questions can you ask when you see this new patient when you're looking at this new thing and figure out what is it? And how do you handle it? And again, I think we come with all that most people don't have. And even big players don't have.

So do you -- are your margins better with small single family practitioners or large hospital practices?

Yes. So I think, again, like any small business, when you're -- when you're a business that did $140 million of revenue last year, the way that net margins overall improve is growing the scale because you've got the cost of being a public company, you've got the cost of developing technology. You've got the cost of sales and marketing and growing over a relatively small revenue base. And so at the end of the day, when you look at the net margin, you say, "Yes, gee, I wish that was a lot higher. On the other hand, Again, we're in this for the long term. We're not at all about how do I maximize my profit margin in 2023. We're all about how do I grow scale? How do I grow margins in the long term. And that's what's ultimately going to give the return to the shareholders. And so there are things that you can do with the small practices who I'd say will be a little bit less cost sensitive. On the other hand, there's some scale that comes -- I mean we have a physical therapy and rehabilitation practice with 2,800 therapists. Yes. Okay. They may pay a slightly lower percentage, but I'm doing work for a lot of claims. And you know what, I'm able to do very nicely there, too. So -- so for us, it's really almost a balanced portfolio. I need to have some in 1 market cap and some and eventually when I put it all together, I've got a good way to get good overall returns.

And how do you market to them, whether it's the practitioners or the large hospital systems?

So I'd say that certainly different because when you think about a solo practitioner, I mean we're not sending a person out to call them. I mean, again, in the last 3 years, people haven't done as much. But I'd say we were having a discussion. I think this was 2017, '18. And at that point, most of our customers were these small practices and there were 90-odd percent of our customers that we never met, never met in the sales marketing process, never met even in the implementation process. We did it all electronically. And that's certainly still true today. When you think about a hospital, that's going to be different. You're probably going to need somebody there pressing the flash talking to them. And frankly, probably leading off more with some of the services that we picked up with our Med SR acquisition 2 years ago. So there, we're starting off with -- we don't have an EHR for inpatient hospital. And we've chosen not to invest and try to sort of outdo the Epics of the world, because we've seen other people spend a lot of money and have very little to show for it and we don't need to go through that learning ourselves. We can learn from their mistakes. So -- but on the other hand, I look at those hospitals and say, they need help in a lot of different ways, whether it's professional IT services, whether it's helping in denial management. I mean there are 100 different ways that we can help them and ways that we can collaborate with the EHR manufacturers who are in that space, and that's the way we approach it.

What are the privacy and regulatory concerns around the personal health records?

Right. I mean clearly, again, whether there's AI or anything else involved, the most important thing is to keep it private to be having people continuously being diligent to make sure that -- I mean there are malicious actors out there. And frankly, with AI, there is more and more sophisticated attacks that happen every day. So we've got to be ahead of the bad guys in making sure that, that data stays private and stays in a way that it can't be -- it can't be accessed. It can't be changed. It can't be a erased. It can't be ransomwared. I mean you need to be watching out for that because that's -- if you don't do that, you've really done a very bad thing for your customers and for their patients. So.

You're not going to have customers.

Sorry?

You're not going to have customers.

You're not going to have customers. You're not going to be in business. I mean -- and we make that a top priority. And again, the fact that we've got a large technical team, and we can afford to do that in part because we can add people on that team overseas who are really skilled who could be looking at this problem 24/7, and I can put more of them to work outsmarting the people with the malicious intent. I could do that better than somebody who doesn't have those resources and doesn't have those capabilities.

Right. Approximately how much does this cost on a per patient basis.

We haven't really thought about it on a per patient basis. I think we've done 1 estimate of what would it -- if I was going to supply a doctor and we picked 1 random doctor doctor #1, who's been with us for 22 years. What would it cost if we deploy generative AI for her to make her more productive. And the number looks like a couple of hundred dollars a month. And then the question is, okay, what are the things we're going to use, what are the costs I'm going to save? And what are the extra capabilities I'm going to give her that would cause her to say, yes, this is a good service. I want this. And again, we're thinking about this -- thinking about it for the long haul. Day 1, you're going to start putting things out there and testing, seeing what's the value and not [nickel and dime] people as you're going through it. But then you'll figure out these are things that are really worth charging for and here's how I charge in a way that makes it worth my while.

Would this, Bill -- would this be like a clinical assistant -- is that the kind of [indiscernible] you're thinking about?

Right. Like a clinical assistant like some -- again, some way that can help this doctor improve the overall wellness of the patient and improve the patient experience. And again, that's the end -- the end goal here is if you're not making the patient better, you're actually not accomplishing that much. So it's got to be focused on that. And therefore, it's got to work through the doctor and through his or her team.

Okay. Backing up to a question -- a follow-up on the privacy. So in your AI models, you need patient data in order to do the AI models. And are you obtaining permission for that data, that patient data?

So if we're going to use any patient data that's identified by a patient, of course, we're going to get permission. Again, there's a lot that you can do with [the] identified data that can help you to understand trends that doesn't violate any PHI concerns. And clearly, that's step 1 is thinking about what can I do that doesn't include that. If I was going to do something that was going to specifically make recommendations for a patient, you probably have to be doing that with their consent if you were using their particular data set.

Right. Okay. And so I think that basically concludes our questions and concludes the webinar. So everyone out there -- actually, before we go, can you confirm the stock symbol to everyone -- yes.

So CareCloud is traded on NASDAQ and the ticker for our common stock is CCLD.

Great. And are you generating revenue?

We are generating revenue back when we went public in 2013, we were generating revenue. And last year, we generated approximately $140 million of revenue and $22 million of EBITDA and positive GAAP net income.

Great. Okay. Thank you. That last question is stuck in there. So thank you, everyone. You have my e-mail address, if there are further questions or further introductions, I can connect you. And before we close out, Bill Korn, and Bill Suthernland, I want just any final words to the audience before we say goodbye.

Just have fun and keep an open mind. And remember, there's a lot of things that are possible that you might have thought weren't, but you might be able to do them today or tomorrow.

Great.

Couldn't say it better. Thanks, everybody.

Okay. Thanks, everyone. Have a great day.

Thanks.