Efficiency, Productivity & Revenue: The AI & Technology Tools Every Biller Needs
In a rapidly evolving industry, staying ahead of the technology curve is essential for medical billing companies. That’s why 4D Global hosted a webinar with top experts to explore the latest advancements in AI, automation, and intelligent workflows. Moderated by Chanie Gluck, the discussion brought together a panel of industry leaders including Sean McSweeney (Voyant Health), Jeff Hillam (Red House Med), Jonathan LaChapelle (Adonis), and Nishant Matthew (4D Global), each offering real-world insights into how AI is transforming the Revenue Cycle Management (RCM) landscape.
The conversation covered a range of hot-topic issues, from where billing companies should begin their AI journey to whether they should build or buy technology solutions. The panel emphasized starting small—such as implementing a denial management dashboard in Excel—and scaling thoughtfully. The panelists established that medical billing companies must act fast to stay relevant in a competitive market.
They also discussed how tools like AI voice agents, agentic workflows, and machine learning are enabling faster, smarter workflows that reduce manual labor and accelerate cash flow for healthcare organizations. As Sean McSweeney noted, “Today’s data becomes tomorrow’s insight, and tomorrow’s insight becomes tomorrow’s data”—underscoring the importance of building a solid foundation for AI success.
Whether you’re a billing company just starting to explore AI or already knee-deep in automation, this webinar offers practical strategies, candid advice, and forward-thinking ideas for leveraging technology in a high-impact way. Be sure to watch the full video, or read the transcript below if you prefer, to learn how industry leaders are using AI and technology tools to enhance operations, streamline workflows, and prepare their teams for the future.
Webinar Transcript
Please note: this transcript was AI-generated and lightly edited for clarity. Minor discrepancies may exist.
Chanie Gluck: Hi everybody. Thanks for joining us and welcome to today’s webinar on efficiency, productivity and revenue. The AI and technology tools every biller needs. I’m Chanie Gluck, the CEO and founder of 4D Global, and I’m thrilled to have you all here as we dive into the exciting world of AI and automation for medical billing companies. Medical billing is evolving at an incredible pace and staying ahead of the latest technology is critical for maximizing efficiency, reducing manual work and increasing profitability. Today we brought together a panel of top industry experts to dive into the must have AI and automation tools that are shaping the future of RCM.
I’m honored to be joined today by an incredible lineup of leaders at the forefront of technology and RCM. Joining me from the 4D global team is Randy, our head of business development, and Nishant Mathew, who is our CTO . Hi Nishant and Randy. Thanks for joining me here today. And thank you to Paul and Piper who are our technical team.
I also have with us our panelists, which are Sean McSweeney. Sean is the CEO of Voyant Health and a national expert in health care revenue cycle management, automation, analytics and AI. He consults for medical billing companies, private equity firms, evaluating and acquiring RCM companies, and in his spare time builds automation and analytics solutions. Sean is also an advisor to several healthcare AI startups which I’m sure we’ll hear more about today. Jeff Hillam is a entrepreneur, investor, speaker and art collector. He’s currently the CEO of Red House Medical Billing. Red House leads with a differentiated value proposition in the marketplace built on tech integrations including AI, RPAs, global mindset and cultural competency, a long standing payer network and patient, relationships. Jeff also has offshore operations, and is heavily leveraged onshore and offshore. And Jeff’s a really good friend and someone that I speak to quite often. And we, you know, have our little brainstorming sessions. So Jeff, thanks for being here with us. And lastly, Jonathan LaChapelle. Jonathan is the director of RCM at Adonis, a revenue intelligence and automation platform built for health care solving operational challenges that impact the integrity of revenue cycle management. He has over 14 years of experience in RCM and medical billing at institutions like Boston Medical Center, Boston Children’s Hospital. Jonathan Adonis is, to me, one of those products that really is changing the way medical billing companies operate. you guys have a lot of cash and a lot of money behind the company. So we’re super excited to hear about what you guys are up to. So thanks for joining us.
Chanie Gluck: Okay, so let’s talk about how technology is reshaping the medical billing world. we know that the, that we’re all being solicited by every single AI company, and medical billing companies are now really turning into tech companies. Do you think that AI or all the pieces of AI RPAs are going to put the medical billing companies out of business? And, Jeff, I’m going to have you start with that since you own a billing company.
Jeff Hillam: Sure. And I, think the answer to start with is yes. There will be a subset of pillars who will fall far enough behind that they will no longer be relevant. There is definitely an imperative to adopt technology. You simply can’t. You can’t rely on what we have, just like you couldn’t rely on a typewriter. Still, there won’t be a place in the industry for people who ignore the imperative to adopt artificial intelligence, robotic process automations, and machine learning. That’s just the way. It’s not even the way of the future, it’s the way of now. So if people haven’t started yet, they’re already behind.
Chanie Gluck: Okay, so if I’m a medical billing company and I’m doing nothing, where would I start? How do I navigate this landscape?
Jeff Hillam: Boy, that is a good question. I think the first thing you would need to do is identify your own internal philosophies toward it and understand where you want to put, some automation. And then you would need to look at your processes one at a time and probably talk to a lot of good friends out there and see what’s available. And then pick one process to start and then work from there. What you know, whether it’s on the front or the back end.
Chanie Gluck: Got it. What Jeff just said is pick a process, go full in on that process, and go on to the next process. That’s what I think. I think Jeff said. I’m going to paraphrase for you, Sean. What do you think? I know you’re involved in a bunch of different tech companies you’re consulting. Where does, where does someone start? give us, give us your perspective.
Sean McSweeney: Well, but I the starting point also kind of ties to where do you want to end up? Because you have to think about where are we now and where do we want to get to and then figure out a path to get there. So I mean even on the question of like are billing companies going to be eliminated? 20 years ago I was in the EMR space and was building an EMR and believed that building companies were going to be obsolete and so they were buggy whips. Two years later, we had started a billing company and decided to go from there. and it’s still hot space but I think yeah, it’s shifting from a service business to a technology enabled service business to what really is going to be a product centric technology business. But most of the industry is just not there. Most of the industry still is body shops throwing people at things, slogging with claims. So if you think in that context of what’s the core service being provided, it’s really collecting revenue, it’s making money for healthcare providers. So whether or not you build things yourself, you sell into the space, you’re sort of an aggregator of these technologies. You build some yourself, you implement from others. You essentially are going to have to move through this technology progression to stay ahead of the space and make sure that you’re performing well, especially relative to the payers who you know, have a lot of AI on their end fighting now. So starting I’ll give one thing so the hp, I did a webinar for them I think August of last year on this particular subject of like how to get started, where to start. and you can pay the HBMA or if you’re not already a member you can go check in our. Again I don’t get any compensation for that. but please check that out. It’s a great way to figure out like where to start. I actually advocate something super narrow like do something like denials management. Super simple down and dirty mvp. Get something in Excel where you’ve extracted data, start analyzing it and go from there.
Chanie Gluck: Okay, okay, well give us a little teaser. What, what did you talk about at the HBMA webinar that we should all give us, give us a little nugget from that. Sure.
Sean McSweeney: So the basic concept was come up with a vision for where you want to end up. and then figure out building blocks that help you get there. And so it could be anything from customer dashboards to a workflow solution to denials, management and plan that out both sort of horizontally and Vertically, where you think, let’s do something super simple, a down and dirty mvp. And the example I give for denials, I’m almost giving away the webinar, but trust me, there’s more value there. if you’re going to do something like denials management, don’t try to build a whole denials management application. Just figure out how to extract the data, parse the data and stick it in Excel and you can use pivot tables and you now have a, denials management analytics solution. I mean it’s incredibly simple, not a lot of time or money. I mean you can do this in like two weeks. And now you have denials management. Sure, there’s way more advanced versions, you can build that out much further and you can build more modules and stream all together and stuff like that. But find something where it’s a pressing need in your organization that has a technology or data component to it and just get started and bang something out. If it takes longer than a couple of weeks, you’re biting off too much.
Chanie Gluck: Got it. Okay, good. Jonathan, let’s, let’s move over to you because I know you have a completely different perspective of this. tell us a little bit about what Adonis is doing that you think is different and setting yourselves apart in this space.
Jonathan LaChapelle: Absolutely. so we always like to think of the world of RCM as kind of like death by a thousand paper cuts. and in doing so we’re finding out there is a lot of these AI solutions, automations, agents, et cetera. You know, all the hot phrases that are out there now. M they solve for a problem, a singular problem. They, they fix point A or point B. The approach we’re taking at Adonis is much more around how do we think about this holistically and create a collaborative, all inclusive agentic approach, whether that be through telephonic, agentic, through RPA based website scrubbing, et cetera. How can we solve for all of the problems, and have that driven by the foundation of our advanced AI, intelligence platform that can identify the issues in a matter of seconds, create the scripting the process and then execute on that through an agent. And what was taking hours for a biller to call on five claims in the past, we can do a thousand of those in a matter of ten minutes. and then provide feedback, solution and resolution in real time.
Chanie Gluck: So I think that’s super cool. let’s speak to the billing company that’s not using Adonis as their platform and they’re using a Different platform. How would they go about leveraging some of the AI tools or RPAs that you’ve built into your product? Like, speak to us through that.
Jonathan LaChapelle: Yeah. So we kind of look at this as a phased approach for a lot of these folks who are currently not on Adonis, but want to get into that agentic solution. the real success behind our agents is driven by the intelligence platform that we have with the ability to create alert and real time feedback. but we’re even in the process now of doing some sort of, like, to Sean’s point earlier, manual review of denials, et cetera, and identifying trends and patterns, for customers who are on the horizon of signing with Intel. But we want to get them up and running, maybe even a little before all the integrations and everything are set up and are in place. So we’re able to offer some agentic approaches for them to show them value early on, and then we can then downstream plug that into our intelligence platform and be a little bit more hands off. but we can still drive it from a manual basis just to get them kicked off and started.
Chanie Gluck: Okay, so that’s super helpful. I want to hear. Nishant, could you explain. Jonathan just talked, about agentic AI, agentic workflows. Could you explain to the person that never heard of agentic AI or what that means? Could you, could you explain this to us in, layman’s terms?
Nishant Mathew: Yeah, I think, I mean, honestly, I mean, if you have an RCM company, you’re utilizing human agents, right. And they are, doing a specific set of task actions. Right. To get to a certain endpoint. Right. So I don’t think that the shift is very much from that. It’s just that now you’re having an AI agent sort of step in and run that process, end to end. Right. And so, so if you think about AI agents, it’s not that different from a human agent. so, you know, conceptually it’s the same. The challenge, I think comes into play is, you know, how do you make sure that these, agents are operating in such a way that they all get to the certain outcomes that you’re. You’re after. And it’s less about getting the task done, it’s more about getting to a certain level of outcome for your customers. and I think that agents, the step forward, honestly, and we’re kind of seeing it right now with, some of the technology that’s coming up with what’s known as mcp, it’s like we are able to now go beyond the agentic model and go beyond, and query external systems and bring that information back into the workflow. So you’re thinking about an entire decision, an autonomy type scenario, that we’re headed into. So, but fundamentally it’s those agents are serving that underlying this needs to be done for us to get to this point. And that’s fundamentally what an AI agent will do. the exciting thing that I think that I also want to tease out to everyone is that there’s another layer that’s being developed now that’s, that’s, that’s providing to be even more exciting than just, you know, agentic workflows.
Chanie Gluck: So. Well, what is that?
Nishant Mathew: Yeah, basically this idea that, you know, there’s a layer that is going to utilize, these agents. But again, we still have a concept of human in the loop where let’s just say that 80% of your calls are going just fine with the AI agents. But let’s just say that you have a payer who does not want to talk to a bot or an agent. The failover is to go to human. Right. and the reason for that is because there’s a decision making process there. Right. So, and there’s a reasoning element to it that a human can do. That’s kind of what people want. The layer that’s most interesting now, that’s being developed now is this idea of doing more autonomous functions and actually get it to not just get this task done but rather close the entire loop because you can have this back and forth conversation with it. AI agent. So again, pretty exciting stuff. a lot to unpack. There is a lot of sort of to just throw it out there. It’s a bit of black box stuff. Like there’s an element of magic to it. and I think that it’s still, yet to be determined what we can see from all this stuff, coming out of the AI world.
Chanie Gluck: Yeah, absolutely. I think it’s very exciting time. I compare this time to when EHRS first came out and everyone was super confused about which EHR to use. And then there’s certain ones that just rose to the top I want to hear about. You know, I’m sure everyone on this call and everybody listening is inundated with emails and LinkedIn messages from every single AI company out there. and we’re all trying to field these calls, right? We’re all trying to like say, hey, is this company legit? Did this company have any users? Like, what’s Going on. and so, and Ashant, I just want you to speak to the concept of using a third party tool. Right. You and I have talked about this a number of times at 4D global. We’ve looked at a tremendous amount of tools. constantly talk to us about that piece of it. using something that’s off the shelf versus building something of your own.
Nishant Mathew: So I think that when it comes to picking a tool, so I, I just want to, I, I just want to acknowledge that it is difficult because there’s a suite of, there’s so many tools out there. Right. one of the fantastic things about this, this new explosion is that the ability to build companies around AI is, is, is moving very fast. Right. So and largely speaking the AI world has been rather generous, in a way. But there’s also sort of like an underlying reason for that, in that they want data. Right. And so when you’re talking about ah, picking a tool for your company, the couple of things that you need to really understand is, you know, what is sort of the strategic partnership look like because in a way you are bringing this person or this tool into your customers ethos. And their ecosystem. And so you want to really understand that the, the data that’s being fed is being fed into this large model. And so the question at the end of the day is then who owns the data? Right. And who, who is, you know, who is primarily going to benefit from all these interactions? So there’s a short term play with tool in the tooling in that we want to stay relevant. Right. we want to maintain that we are not simply just a human scaling machine, but rather we’re bringing in technology to optimize as much as possible. So the short term strategy could be that I just want to remain relevant and tell my customers that we are utilizing technology, we are optimizing and all that. But I would encourage to Jeff’s earlier point is that what is your philosophy? Where are you going with all of this? Right. And what is the relevancy in 5 years time when all this data has been fed into all these AI tooling? What is the impact of that? So when it comes to picking up tools, make sure that you are also considering the fact that you are currently in the middle. And that may not be the case in five years time. and so there’s a direct relationship between the AI company and your customers.
Chanie Gluck: I mean, Sean, how does that resonate with you, with what Nishant said? I know you’re involved in a bunch of AI companies. How do you see this, ah, this piece of it, the data is. Is really quite valuable.
Sean McSweeney: Well, I mean, for many years people have been saying, like, data is the new oil. And so, you know, even five years ago, I was running into situations where, you know, we were trying to do something call it, you know, benchmarking kind of solutions where we were doing, you know, MVPs and rolling it out. People wanted us to pay them to benchmark for them. I was like, you’re out of your mind. Sorry. Your data individually isn’t that valuable. but yeah, so I think there’s a problem in the sense that there isn’t an easy mechanism to value data. but absolutely, we all know that it has value because there isn’t sort of a commercial market for that outside of something like pharma. but the subject that Nishant brought up of who owns data is important, not even because of sort of ownership, but in terms of, like, how the data is going to be utilized and the value that’s derived from that. And so I just had a demo this week from a call it a analytics solution company, who said that they have a machine learning portion, of their business and they were prioritizing claims based on, you know, probability of success. And the question I asked was, okay, you’ve got 500 data sets. M. Are you training off of the 500 data set superset or is this just our own data? and they said, it’s just our own data. I said, well, that’s not super valuable because you may just be reinforcing bad behavior. You know, we kind of want to see, you know, what’s. What’s the learning basis out of a larger set of data. and so if companies can figure out how to do that, I think there’s a tremendous amount of value. That’s one of the things we were trying to do some years ago. you know, under that prior company, Apache Health, you know, we ran into all these kind of problems where people didn’t want to give us their data. you know, even when we’re offering to do all kinds of work for them for free. And it just, it was very challenging to deal with. So.
Chanie Gluck: Okay, so thank you. And so, Jeff, from a medical Billing company perspective and speaking to other billing companies. Is this something that you think billing companies should be concerned with? Like, what’s your. What’s your position on this?
Jeff Hillam: well, I think they have, a lot to think about. I think that when it. So I’ve talked to a lot of billers, and there is, I think, a lot of people who are concerned out there. My feeling is that. That the best thing that people can do, even if they don’t to. To try to remediate their own concern, as it were. The best thing they can do, even if they’re not ready to act, is to research, and just be constantly researching. We’ve already. I mean, we’ve talked about a few things already, and there’s a few really good, specific things we’ve talked about. But if people aren’t ready to make the investment, if people aren’t ready to pull the trigger in one direction or another, if they’re weighing down with analysis paralysis, like, Nishant you brought up there.
Sean McSweeney: Just to answer that question. Should building companies be concerned about ownership of the data? The short answer is no. And the reason why I say that is this data is highly perishable, especially in this industry. And so the, you know, data from last year has relative, I mean, incredibly rapid depreciation on the value of that data. So if you’re sharing data and you somehow think you’re giving away the farm, it’s not worth anything 14 months from now. So I really wouldn’t be concerned about that as an obstacle to moving forward, especially in the short term, five years from now, you know, we can have an interesting conversation right around that.
Jeff Hillam: I think there. There is one thing, though, there, that the data is sellable in a lot of ways. even if it’s old, there’s a lot of people that are outside of our space who are trying to capture our data. People who, if I’m billing for urgent care, is people who want to know how many vaccine shots are going in arms and when and whatever. So even aging data has a, marketplace in our systems.
Chanie Gluck: Jonathan, how do you see this as being an issue or not? Like, what’s your perspective on this?
Jonathan LaChapelle: so, I mean, from. From our perspective, in our world, the data lives with the customer, right? Like, we. We integrate into their platforms. We just feed that into our system. but from an intelligence perspective in how our platform continues to learn and grow, we Absolutely. Rely on some historical data. Because when you buy a platform like our intelligence, we don’t want to just say, okay, you started today. We look back a year, two years into your historical data to start identifying trends and patterns. So our machine learning, our AI can kind of do its thing and understand what were your previous hurdles that you were trying to fight through and that we can build off of. so there definitely is a value. And if I leave the Adonis space for a second and go back to my Boston children’s space and those worlds. Ah, it’s a. Jeff’s point. And Sean, we started getting into the epic cosmos world where we were looking at stripped clinical data to make really important research decisions for our patients based off medicines they took and diagnosis they had later in life. there’s always the need there. And there is a level of value, I think, in historical and current data. Being married together.
Jeff Hillam: If I could maybe just say one more thing too, that sort of goes with that, and, and that is that medical billing companies shouldn’t undervalue their data. Though I, I do think it’s a little bit scary to give your data up because once it’s gone, it’s gone. You know, once you’ve given it to somebody, you don’t get it back. but there’s, there’s a lot of people willing to pay for it. And billers are not really used to negotiating prices for their data. That’s not part of their revenue streams. But it can be a revenue stream, you know, whether you’re planning to use the tool or not. And participating in that process, I think is going to become an important part of how it all develops. Because one of the biggest problems with the medical billing AI phenomenon is, just the permutation count, of, of. Of workflows through the medical billing. You know, there’s, there’s just hundreds of millions of, you know, this to that to the. There’s just so many different permutations that the more data can be, participating in the development ecosystem, the better.
Chanie Gluck: Yeah, Jeff, to your point, I think, on the resolution, like, how do we resolve this particular claim and the workflow that goes along with it. To me, that’s the value, in the decision making and getting the machine to be smarter and how to resolve these claims. Okay, I just want to say this, everybody. we want this to be interactive with our people. It’s amazing how many people we have on the call. I’m just blown away by how many of you are interested in this topic, so fantastic. but if you have any questions or if there’s anything specific you want to ask a certain panelist or just a discussion topic, please drop that in the chat. We want to hear from you. So go ahead and drop that in and we’ll leave, some time at the end for Q and A. You wanted to say something?
Nishant Mathew: Yeah, I just want to. So I agree with everything that has been said so far. I think the thing that, I think we tend to think of data as probably the most raw, like the most bottom layer of data. And I think today’s data is tomorrow’s insights, which becomes tomorrow’s data. And I think the thing that you don’t want to be in a situation where you’re thinking that that original raw data is where the value is. No, what’s happening is that we’re building layer of insights and then on top of that we’re building another layer of ins and each of that becomes the next level of data. So the thing that I want to encourage folks is to be like, what are you doing to capture insights into what’s happening with your customers, with your business? And if you’re not in that process of developing that insight over insight, to go to Sean’s question. Yes, data is perishable and in some cases it’s only good for that, for that interval of time. You know, its value is very short lived, but it’s the insight that you gather from that short interval of time that’s the, that’s the next layer for, for the future. So keep thinking in those terms and not just sticking to that. Well, we have the fundamental data and that’s all we really need. No, it’s like building on top of that. Today’s data is, is tomorrow’s insight and tomorrow’s insight is, you know, tomorrow’s data. So that’s how it gets built out.
Chanie Gluck: Absolutely. Okay, let’s dive into. I just want to say this first of all. a few of us started this tech forum that we meet every month and we look at all the different, Jeff’s part of our group, Sean’s in our group, and a few other people that are really using, data tech a lot and trying a lot of different things. So, if anyone’s interested in getting more information on our forum, please reach out to me and let me know. we’re only looking for people that are in this space and using, different things like we are. but I want to talk about specific products. specific doesn’t have to be a specific company, but a specific part of the process where you would recommend that people start looking into this area. and I’m going to kick it off with, ah, Sean, is there something, is, are there some products that, that you’re into and you got to give the disclaimer if you have ownership interest, company. Okay, absolutely.
Sean McSweeney: I, I fully support that position. Well, I mean, I think in the broader context of technology, if we look at some different areas like machine learning or large language models, or even some of the agentic AI or AI agents, you know, large language models were super hot for the last couple years. AI agents or agentic AI really has sort of gotten, starting to get the attention much more this year. But I mean, backing up to like conversational large language models. You know, if we look practically, what’s really AI as opposed to, you know, automation tools and you know, analytics and other things. And again, I’m a huge believer in ML and don’t think it’s utilized enough in this industry. And this is a really a data centric industry. But talking about what’s hot, you know, conversational AI where you can, for example, have an application that actually interacts with an insurance company or even with patients. and again, full disclosure, I’m a consultant to several of these, one of which actually has started to get to significant scale. Super dial. And the concept there is you have an agent effectively that is doing outbound calling to insurance companies to do things like pre authorizations or check eligibility and other things. And you know, the data that’s coming back and not just specific to them, sure they’ve got good results, but there’s other companies like this as well. The data coming back is that they’re very effective. The, you know, relatively small number of those calls, for example, have to roll over to a human or are not successful, again, depending on how you have the workflow set up. And so those are having real significant impacts where you know, what used to be fleets of people, 50, 100 people doing outbound calls is being radically chopped. and again, all the data isn’t in, in terms of whether you, you know, have 80% of them replaced or 97. And some of that depends on the complexity of the workflow. But those are actually, I’m seeing really, really positive return on results, return on investments, for those types of applications. So. And I’ve got some other things, but I don’t want to just dominate the call so I can throw some things out later.
Chanie Gluck: Okay. What about you, Jeff?
Jeff Hillam: Okay, so when I’m thinking and sorry, I’m actually just trying to get out of the sun because I am getting blinded. So let me just rearrange there. Okay, thank you for that. So I am thinking of 2025 and I am thinking of ways that people can look at the softwares to bring in and I’m thinking of ways that people are changing how they think about integrating AI. I, I do think, yeah, I do think the, I agree with you, Sean. I think the agentic conversation has got a lot of a lot of heat underneath it right now. People are really interested in it. I think a lot of people struggle to generally with knowing, you know, and committing to one software. And what I think if I, if I were just to maybe mention a trend that I think is at least worth noting in a group that’s having this conversation is that a lot of billers, us, Red House included, we do in a lot of ways just rely on the, the PM and the, the EMR systems that we’re functioning within. And those platforms are having an evolution in their philosophies also and they’re not all going in the same direction. And knowing which PM and which EMR and I, I, boy, I do think that those need both be talked about together. it, it’s important to make sure that you’re using a system that’s going to align with your philosophies because whether you’re using a system that’s going to allow you to do backend, API, whether or not you want somebody that’s developing it on their own so that you’ve got a system built on, you know, new coding and not something with old bones trying to put new carpet in. you know, it, it’s definitely frustrating and I know people probably might feel this out there in the billing world where they feel like they’ve got like a Frankenstein’s monster of tools plugged in and boy it can be hard to, to manage that and have them communicate. And, and those types of scenarios are going to continue to, to plague you know, all the billing companies trying to build up their AI capabilities. And so it’s, you know, it’s worth noting that some of the groups have open APIs, some are really interested, some of them just have an app store where they say anybody can get certified and they can just plug in and you can go buy. And we’re never going to develop an AI anything. And so it does matter which base platforms that you use. For your client services. and then the one more thought is, is I feel like an another under discussed piece of the AI implementation. And I’m not going to try to get outside and say, oh, you should do your marketing. I’m staying, I am staying right on our operational value, prop as RCM, folks. But it is to look up and down the value chain a little bit and understand what areas of you know, look behind you, look in front of you in the process and say what else can be automated to make this part of the value chain better? You know, I, I was listening to one of you talk a minute ago and I, I was reminded of a software I did a demo with a few weeks ago of somebody that says they take the billing data and then they stick it into the EMR so that when a doctor is trying to close a chart in a way that always gets denied, it won’t let the doctor close it anymore. And I thought, boy, that would make the billing easy if the doctor was just like, you know, try to sign it and close it. And they’re like every time you do this it denies. You can pick one of these three options. Each one of these options was paid in like in a remediation process. And so there’s really cool things to do looking back and forward in the value chain of RCM operations that it is being just being included. And I think these are really fun parts of it that can help companies differentiate instead of just, you know, it can put you in a position to act and not just be acted upon.
Chanie Gluck: Okay, so Jeff, you made a good point about the PM systems. Do you have any PM systems that you like or you feel are investing in the AI? I know of course Adonis is doing that. We know that they’re like highly motivated. anyone else, Jeff, that you see is doing that?
Jeff Hillam: Here’s I, I without passing judgment on any of them, I think for me I, the systems are in, they have their pieces that are better or worse or whatever and everybody has their own opinion. So I’ll, I’ll, I’ll hold back my opinions. But for example, somebody like Veridigm, they’re the ones by the way with just an immense app store. I don’t see them really doing a lot of internal R and D, but boy, you can plug in 200 different apps right into the system. I know that groups, like Tebra have tried to reduce the barriers to entry to be able to have vendors come in. Although I have noticed that a, few of the softwares, including AMD, have, Tebra, AMD, a few others even have started charging vendors, a rate to be able to participate on their back end in some instances if it starts getting big. ECW has done a good job, I think, of, of, of making internal changes. Boy, but the thing that I think is missing in a lot of them is as they develop or as they start plugging tools in, the reporting is not keeping up. And so the old reporting that is in the old systems is just no longer valuable reporting because a lot of the systems are relying on outside vendors to make their system better. They’re not always in control of the analytical experience. And more so I, I think that EMR companies are at a little bit of a crossroads right now, and in the next couple years we’re going to see them having to make some hard decisions about what their future looks like. And so my advice always is to keep your finger on the pulse and don’t stop doing demos. Don’t be complacent. You know, just constantly be ready to pivot in case you see that, you know, your horse isn’t going to win this race and be ready to go to the, you know, you ready to hop from one to the next.
Chanie Gluck: Well, we have a ton of questions flying in, so I really want to make sure we can get to this. And I’m going to try to do my best. but I want to ask this question. Dana asked. Does it make more sense for billing companies to purchase AI solutions directly from the AI technology company or purchase them through a wrapper company that offers or manages an interface between client and AI tech and human in the loop services. So I guess each one of you probably have your own perspective. Like, are we relying on the PM system? Jonathan, in your case, of course, yes. but I don’t want to put words in your mouth, Jonathan, so please, if you disagree with me, tell me. but. Or should they go to a third party? So, so how, and I guess, you know, Sean and Jeff, you could probably answer this. the best. if you’re working with these systems.
Sean McSweeney: Sure. I’ll go, since Jeff, you just went, but I’d love to hear his perspective as well. This kind of comes back to that subject of not just buy versus build, but are you going to be sort of an aggregator of different technologies? And so I think Jonathan has espoused, and it’s not good or bad, sort of a comprehensive approach to everything where we kind of want to build everything. and if you have the capital to do that, fantastic. If you have scale, you know, I’ve worked with building companies that have half billion dollar valuation, they’re private equity backed. That’s a very different kind of scenario I think than companies that have 5, 10, even 15 million revenue. But at the very least, you know, you should be thinking in terms of like we have all these different things that we need from a technology standpoint. Each one of them you could think of in terms of a buy build kind of thought process. But then you have to think comprehensively how are we going to string all these things together and make them work? But I, you know, my belief is just watching this industry over the course of the last couple of decades depending upon a single primary system, you know, like pick somebody, ECW type or whatever it is, and we can talk about the various benefits of each of them individually. There’s just no way they’re going to do everything well, much less invest in all of these things. Especially since most of them are at root EHR companies where they derive the revenue primarily from physicians deciding, deciding to use that application for ehr, which means they decide on a clinical workflow, not on revenue cycle management or data analytics or technology or other things. So they know what butters their bread and that’s really that portion. So it’s unlikely. And so to be tethered to one, I think is not the best plan in terms of risk mitigation. and if you think about the fact that you’re likely going to have to use more than one when clients require you to log into the system, at least that’s the status quo in the industry. I have a different view of centralized billing and interface to all of them. But assuming, you have to do that again, you’re outside of that control. So we’ve espoused frequently having a data warehouse. We extract data from multiple systems and you can start to build applications on top of that. Everything from analytics to automation to artificial intelligence solutions. Again, whether it’s running your own ML, because again you can get somebody to do some ML for you on your own data set. Relatively easy, there’s a, a ton of people out there. or you can just plug in third party applications if you have that, and then you have much more flexibility going forward of hey now, do we want to build our own rather than buy that, or do we want to switch vendor A for vendor B and go from there? So that, that’s kind of our philosophy.
Chanie Gluck: Jeff, what about you?
Jeff Hillam: I, I actually like that, I would not, I would not differ too far. I think the biggest con of relying on an EMRPM system. Sean, I love, I love already the way you described that, with them, yeah, it’s, you’re relying on their investment, you’re relying on their tech roadmap, for the future of your business, which by the way, if you’re a billing company, your core competency, your main value prop to the clientele universe you have, is that process. So hitching yourself to a wagon where you’re not in control of the development is a risky proposition. Though on the flip side, when you start going over to some of these other groups, there is a lot of siloing and then you end up, you know, a little bit out of control of it anyway because the groups don’t have the ability to work in every system you need to work in. And now you’re creating fragmentation and your ROI needs to start considering the complications of, you know, training and logistics versus, you know, versus ease and ROI and money. And it, there, there are cons on both sides. And yeah, I think a little bit of hedging and having a little bit of diversity and experience and keeping track of it is good. And boy, not to, not to jump the gun if this is going to come. But I think the other thing is, is that making sure as you use the different tools, you don’t silo the tech skill set within your bill co. You can’t just have your managers know how to do it and your, you know, your, your, your billers, your, your teams, they’ve got to, you’ve somehow got to know what’s going on. You got to be ready for them and start training them now to be bot trainers going forward because they’re the ones that see the process day in, day out. And so regardless of who you hit your wagon to, it’s going to be your team that’s going to make it work. And so you got to start looking at your people as people who help you with technology and not just people who like, smash a keyboard.
Chanie Gluck: So I’m going to say here, shameless, plug for 4D Global, what we’re doing is our subject matter experts are the people that are creating the agenda workflows and they’re the ones that are saying, hey, this is what needs to happen with this claim and we run it through this, agenda workflow. But that really, you know, depends on every, every client’s a bit different. Every client has a different payer mix, different denials different software. And so if, you know, if the volume is there and you have a good partner, many times, you know, we’re so immersed in it, we’re living, breathing it every day. so that’s another option. okay Dana, I hope I answered your question. let’s knock out a few more here. and Dana said, very insightful. Thank you.
Chanie Gluck: How can a company with no prior AI experience ensure data quality and integration when using an off the shelf AI to handle billing and claims? And what initial steps should they take to prepare? Jonathan, why don’t I give you that question?
Jonathan LaChapelle: So there, that’s a, there’s a lot there. So this is one where I’m just trying to think through all the pieces before I, before I speak on it. So when you think of off the shelf AI sort of services and like out of the box solutions, to go back to what I think Jeff and Sean were talking about earlier is the PME EMR system and their ability to coexist with these AI solutions and understanding the integration point, understanding their strategy. You know I can say, and I won’t share names of different EHRs and PM, but some play really well with others and some don’t. and that creates a very difficult path for us in some scenarios we’re fortunate where a lot of the heavy players in the game. I’ll use one. Athena Health for instance, we work very closely with is it a valued AI partner, allows us to kind of never be just an out of the off the shelf, out of the box solution. We’re very customized to the customer need and I think that actually is what separates us from a lot of people is to your point earlier to Chanie like you use your 4D resources to help drive these agent approaches and another shameless plug for you. We help. We also use 4D resources to help us with some of this as we identify trends and we leverage your resources for some of our billing, manage billing. and we do very much so, kind of always take this iterative approach of expanding and growing with the customer need. and it’s never really just off the shelf, it’s never out of the box. It’s here’s our core offering. But then we get to know you, we get to work with you and then we go down this very intense path of customization and building for the need and not just saying this is what we offer, this is what you get. we’re very much a solution for your problem across the board.
Chanie Gluck: Just which is the right approach, right? There are so many nuances in our industry. Okay, so thank you for that.
Jeff: Can I add one thing to that?
Sean McSweeney: Can I add one thing to that? Just because it took sort of data integrity, whether it’s an off the shelf solution that’s trying to plug into your practice management HR or you’re building yourself something yourself or whatever it is I’ve seen with a lot of these systems and I don’t want to name bad players, but when, whether you’re accessing it via an API or other things, there’s often data integrity issues within the core practice management HR system. And so and there even may be data structure kind of issues and other kind of challenges. I mean I’ve literally seen where you do a call and the return comes back with information that is different than what’s in the billing system. Like the patient data service is different. The insurance company primary insurance, things that make no sense at all. There’s clearly a data problem in that system. And so you know, this process isn’t specific to whether it’s an application or anything else. Is I don’t think enough attention in this industry is focused on like do the reports tie out and match up. I mean is there actually data integrity within our system, much less, you know, to utilize in these other functions. So I would be wary of your own system and also even just how it structures the data because it can do all kinds of weird anomalous things.
Jonathan LaChapelle: And I’ll just build off that again really quick. Sean, you’re absolutely right and that’s something that I, not to plug ourselves too much but we, we’ve done really well with is we don’t just have like one format that you have to match to work with us. if you look at our report structure and even the fields that will report out on it does vary based on the PM or EMR system that you’re utilizing. And our Athena customers data output looks very different than those of our AMD or ECW customers, and will match the need. but to your point, I think historically we found in this world that it’s very like you have to go one to one in your mapping or we’re not really going to be able to function properly here. and that’s one thing I’m very proud of. What we do is we don’t require that one to one mapping in every scenario and we can be nimble and flexible in how we ingest and replicate your data.
Sean McSweeney: And I believe that you are agree to this. I’m not I’m saying, yeah, weary of some other vendors. Again, I’m not trying to plug anybody and again we have no relationship to you directly, but there are some where you have to really be careful about this. So I, but I can hear from everything you’re saying that you guys got this dialed in.
Jonathan LaChapelle: Yeah, no, absolutely. And I’ve been part of epic, EPIC implementations and Athena implementations and everything across the board at a number of institutions. And mapping exercises are tedious and take a lot of energy and resources to make sure that that data comes out clean and goes in clean. We always said bad data in, bad data out. And that’s the crux and the failing of a lot of places that have issues there.
Chanie Gluck: I have another question from someone that wants to know what are the best sources of information to stay abreast on? AI tools, quality and efficacy? Anybody have any resources they want to share? I know HBMA, MGMA, newsletters.
Jeff Hillam: Yeah, I was going to say our tech forum is developing and ah, you know Chanie, you were the insightful one there. It’s a great question and I think the reason why our tech forum comes up at all is because there is not really a great spot for that. HBMA does have a lot, MGMA does have a lot, but sort of built, you know, for the billers, by the billers. This space is a little quiet. So you know, we are working to try to gather data to create some sort of hub of information and you know, let’s hope that that really develops and blossoms and that we can bring some value to you guys and, and help the industry grow.
Chanie Gluck: Yeah, I’m super excited when we notice this void of you know, no one’s really talking about it, no one’s taking the thought leadership. It was a great way to get some really smart people together. So I’m super excited. if you want to know more. Well, we’re going to post a poll at the end of this just to get some of your feedback and see what other topics you want to hear about. I still have a bunch more questions that but I want us to be mindful of the time.
Sean McSweeney: Can I do one more, can I do a plug, just a personal plug in that. Which is, if you, if you check our blog, we have some articles about these subjects specifically in the context of RCM, broadly technology, rc, you know, automation and you’ll even Find that I do take a contrarian approach on a bunch of these things where I basically almost crap on a lot of the hype around some of these technologies. and I do. This is not specific to this question, but I do. If I’m going to make one point here relative to all the hype around AI, whether it’s agentic or large language models, most of this industry doesn’t have the core data infrastructure in place where they’re extracting data to be able to utilize it effectively for AI. And soo I think thinking in that context of get your data, manipulate your data, transform it, centralize it across systems, and make sure you have data integrity will give you so many options in terms of what to do with it, whether it’s automation or other AI tools. But that’s not happening in a lot of companies. And so it’s almost like we’re missing the forest for the trees. We’re all excited about all these AI tools and then nobody does anything because it’s not easily implemented in their organization.
Chanie Gluck: I would say from my perspective uniform systems are helpful. When you have a bunch of systems it makes it more difficult. That’s what I heard. Number one. I also think that if you have your Carson rocks codes like listed and that makes things way easier and them being accurate I think definitely helps in creating these workflows. What else do you think we should be doing if any, if, if there’s anything else to not to basically look at what we could immediately solve. aside from that, is there anything else you would look at other than those two things?
Jonathan LaChappelle: so I’ll, I’ll jump in and I’ll just say one of the things we see as we get into this is running a clean shop on like contracts, credentialing, your, your electronic remittance, all set up those enrollments. because a lot of this information relies on accurate remittance coming back into the platform and allowing us to analyze that and Even like the two way communication of the 276s, the 277s, like all those information coming and going, that’s the, sometimes the best pulse check, you know. Then you can rely on the second layer of looking through the clearinghouses, looking through the portals, even using the agents to validate some of the claim statuses, et cetera. But getting like a really clean shop on making sure all your enrollments are up to feed. Your providers are credentialed as you think they are and just going through all those little nuances in making sure that you have a really clean shop before you start going down an AI path and relying on somebody else to feed that to you, make sure the data is good for them.
Chanie Gluck: 100% credentialing is critical. I have another question. Go ahead, Jeff.
Jeff Hillam: I was going to say, I think as you look at it, I love that point about keeping a clean credentialing shop. That was a, great point. I think there are, you know, we sort of have a few guiding principles. I don’t know that this is all of them, but I think that in the end there’s a few things that you start looking, boy, there’s some cool tools out there. And you, you know, you end up like you’re walking through a little gadget shop and it’s sort of fun and you get excited about it. And sometimes just because it works doesn’t mean it’s better. And for us, there’s always the idea that we need to make sure that we keep our eye on the prize a little bit. And for me, the first thing to ask is, does this tool have a long term chance of being a good move on the chessboard against the insurance companies? Because if it’s just going to get countered, it’s not going to. It’s. You’re spinning your wheels.
The next one is, is this actually focused on velocity of capital? Because if it’s not, and you’re just fixing a process, you’re not actually going to make money and your ROI is going to spread too far. And then the last one is, just because it’s automating something doesn’t mean it’s not screwing something else up. You know, we’re all floating down the river in a canoe, and if the, the guy in the back is, you know, not paddling at the same time, or the person in the front is, you know, whatever you do when the, when the front moves, the back moves. And so you just can’t say, oh, I’m going to automate this and assume it gets better because it doesn’t mean it doesn’t mess something up down the road. And so you just got to make sure you got your eyes up.
Chanie Gluck: Okay, so what I’m hearing from all of you is that there still is a human component to all of this and we can’t completely do away with humans. Am I right? Or where do we stand and where do we see this going four to five Years from now, like we know where we are today. What does the future look like? Jonathan?
JJonathan LaChapelle: We’ll never in my opinion fully replace the human element as the, the evolving world like with the payers in particular and their defenses against what we’re doing continue. We’re going to continue to need to learn and grow, identify these trends and patterns. Yeah, we can do that through our, like through intelligence platforms and AI tools there. But there’s always going to be the need to do the further research, the digging, going to the granular claim level to understand what’s happening, and to ultimately feed what the machine is learning, what the automation process should look like and what you’re trying to solve for. So in my opinion, I think you know, four or five years from now we’re just going to be solving for a different set of problems that the Payers have put in front of us to solve for then. I mean we’re, you know, we’re looking even to the same this point now of using bots and sort of AI for getting policies into our platform to help build the logic around. Okay, this plan type requires these certain pieces for pre certs to get the claim out the door cleanly. But we are only subject to what’s actively in the policy. If there’s a new change coming down the pipeline, it’s in the newsletter. You know, that’s another level that we need to be aware of and try to plan for. And there’s just a million pieces that we’re always going to have to have the human eyes on things. you know, that may be less and less but it’s never going to go away. There’s always going to be a growing developing need.
Jeff Hillam: Yeah, I feel like boys, this is a fun place to ideate. yes. And you wonder where the balance is. And sometimes I wonder if that’s one of those. There’s no kind of answer. There’s no balance. But I do think in a lot of instances that the theme is driven by fear. and you know, we, here we all are sitting on our computer. I have absolutely no whale oil in my house and I’m absolutely fine. And I don’t miss the fact that nobody’s working in that industry anymore. it’s just going to change and to, you know, I, I’m a big fan of you know, Clayton Christensen, you know, former Harvard professors. Creative destruction right there is this idea that we just gotta, we gotta tear it down to build it back up. And that’s going to apply to our workforce also. And that is just something that needs to be in everybody’s strategic plan. If you don’t have an HR and a people development and, you know, what do my people need five years from now? Because I don’t know if any of you are parents out there. My philosophy has always been whatever my kid needs five years from now, I need to be teaching them now. and my staff, in a lot of ways, it’s the same way, if they’re going to need to know how to use AI five years from now, well, they need to get to learn to use it now so they can, you know, crawl, walk, run right along with our corporate development.
Chanie Gluck: Absolutely. This has been such a fun conversation, guys. I truly enjoyed it. I felt like we can go for another two hours. Maybe we’ll do a part two. I have questions here about coding and operative notes and how to use them. I think we could do a whole hour on all the different coding stuff that’s available out there. And we’re developing this whole list of a bunch of different products as well. we’re going to send a survey at the end of this, you’ll get an email. Please communicate with us. Please let us know. Randy will read. Randy, thank you for being on. Randy will reach out and follow up with some of you that have not gotten a chance to respond to man. Moderating this group of, this amazing group of really smart people, has been, has been a challenge here for me. So hopefully, we’ll be able to get through more of your questions next time. Again, thank you so much. And follow us and we’ll continue to provide information and keep you abreast of everything that’s going on in our industry. These are super exciting times. Really. These are really exciting times. I wouldn’t be afraid. I’d be excited. it’s just a really great, sector to be disrupted and to see innovation and new exciting things coming on board. So stay tuned, for more of this. And to our panelists, Jonathan, Jeff and Sean, you are just a wealth of information and a pleasure to have. Nishant, thank you for your, for your, insight as always. Randy, Piper, Paul, appreciate all of you. Thank you so much and have a blessed day.