Throwing The AI Party Healthcare Deserves w/ John Williamson

Erik Sunset: [00:00:00] All right. Hello and welcome back. I’m Erik Sunset, your host of the DocBuddy Journal. Here at DocBuddy, we deliver healthcare solutions that take the pain and cost out of broken workflows like with OpNote, which gives ASCs and their affiliated clinics the power of instantly generated operative reports approved from the point of care.

Erik Sunset: You can learn more about Op Note and all of our solutions at docbuddy.com. And today we’ve got a really great guest. We’re joined by John Williamson. John is the founder and CEO of Lockbox AI. John, thanks for joining us.

John Williamson: Yeah, Erik. Thanks a lot. I really appreciate it.

Erik Sunset: Yeah, it’s our pleasure to have you here. Glad you could make a little time for the DocBuddy Journal.

Erik Sunset: And before we get into the, uh, the meat and potatoes of the episode, what else should our listeners know about you and your work at lockbox AI?

John Williamson: Sure. Yeah. Uh, well, I’ve been in health care now for 12 years, starting clearing house layer and then building. A couple of different automation companies. My, my latest called Lockbox AI. Uh, we basically sit on top of EHRs, clearinghouses, payer portals, and [00:01:00] implement, uh, analytics, workflow, and automation between those entities, uh, for the purpose of RCM teams, often at EHRs that are.

John Williamson: Doing RCM at scale.

Erik Sunset: Well, let’s, let’s scratch at that a little bit before we get into the meat and potatoes. We were joking before we, uh, we started to record that the, you can’t wait for your EHR vendor to throw the party that you want to have, whether that’s. Like with the functionality lockbox AI delivers, or like with our DocBuddy’s workflow solutions.

Erik Sunset: Um, I’m not shy to throw spears at the space. I’m gonna call right back to those best in class scores. You know, your best in class leader is hovering around a B minus. You know, you may get lucky in a year and be solidly at a B. That’s a big problem for the, for the industry, for healthcare at large, as well as for providers that need to use these softwares.

Erik Sunset: Their care teams, their administrative staff. Why do you think that is, why have EHRs not blossomed into the technology that they could have been all along?[00:02:00]

John Williamson: Oh, you’re gonna, you’re gonna get me on a rant real quick. So, uh, EHRs are frustrating. So the story really begins. If you remember back to, uh, the financial crisis and coming out of the financial crisis was, uh, this giant spending package. I think it was like Sub 1 trillion, so it doesn’t even count anymore.

John Williamson: But at the time it was big and there was a big chunk I think 32 billion or some number like that that was put to Literally pay physicians for their practice to adopt an EHR. So in the span of about three years They had this program called meaningful use. There’s some great commentary out there about who designed the meaningful use feature set happened to be, uh, you know, the owner of the Death Star.

John Williamson: I mean, epic, uh, that designed the features that they already had that were required. So all these. Many EHR vendors, some of which had been around a while, some of which were new, threw a bunch of features at the wall to sort of hit this meaningful use deadline and so that [00:03:00] their positions could get paid for using an EHR.

John Williamson: And so this approach to essentially bribe the provider market into using EHRs led to a whole bunch of EHRs sort of having essentially artificial demand, and then threw a bunch of features at the wall, which then got followed up by private equity buying most of the EHRs. And, and the curse of private equity, of course, not always the case, uh, but they buy something and then they sit on it.

John Williamson: It’s in maintenance mode. So just about every single EHR out there is in some form of maintenance mode. And that’s why the interfaces look like they do. That’s why they’re slow. And you log in through RDP and all this other stuff. They made it so hard. And so there’s nobody to bring innovation. I mean, heck, try and get a good report out of any HR or design one or have visualizations that do anything that respond.

John Williamson: I mean, it’s the worst. You know, if you look at an EHR is equivalent to an ERP [00:04:00] system, you know, SAP and, and these different systems, they are the least useful. And unfortunately, because so much of the data flows through them, they’re impossible to leave some of the work we do. Is to help migrate from other EHRs to our partner EHRs.

John Williamson: And you can get maybe 50 or 60 percent of the data through sort of easy mechanisms, but the long tail of content is difficult. So we have to write automation to pull data out. And so no one can leave their bad EHR. So they’re stuck. And so we have hundreds, I think there’s over 500 in the long tail of EHRs out there, most of which are in maintenance mode.

John Williamson: Most of which people can’t leave because too much of their data is wrapped up and in their business operation is wrapped up in it. And nobody’s doing anything about it because they’re just cash cows now.

Erik Sunset: Yeah, you hit the nail on the head. By my last count, and this is all through that, you referenced the American [00:05:00] Reinvestment and Recovery Act of 2009, which had the HITECH Act in it, which was the Meaningful Use Provision for both an incentive and a penalty to get one of these EHRs. I remember in the early days of Meaningful Use, the expectation was that we would consolidate down to maybe a couple dozen different EHRs.

Erik Sunset: You’ll have the big ones that handle multi specialties, you’ll have a few specialty specifics, maybe for your ortho clinic, maybe for OBG, maybe for pediatrics. By my last count, John, there was still just about 700 certified EHRs on the market. And that’s a, that’s a far cry from the 25 or 30, you know, we thought we’d end up with uh, 15 years ago.

Erik Sunset: And to your point, they’ve just totally stagnated. They’re not something providers want to use. I do want to be clear as a patient, as a health care consumer, it’s probably a good thing that there’s an automated drug to drug, drug to allergy check that happens. So there’s some good that came from EHRs. But if you ask a provider organization, it’s just the bane of [00:06:00] their existence.

John Williamson: And it’s so sad, too, because, you know, I work on the revenue cycle side. So we’re looking constantly. We’re narrowly focused on what’s going wrong in the revenue cycle. No one cares what’s going right only about what’s going wrong. And it’s. These data, these processing errors, sort of things that could have been done differently that resulted in a denial or things that fell through the cracks that ultimately get written off.

John Williamson: Uh, so much nonsense, like the financial impact of having a substandard E. H. R. Is I don’t even know if you could do the math. It’s so high. If you reimagine the EHR of the future, you wouldn’t end up with so many preventable issues going on. I mean, they just but nobody’s going to change. They can’t. I don’t think they really calculated what the switching costs are.

Erik Sunset: No, and you know, early on, you’d see quite a bit of switching. Somebody would buy one, implement it, and be like, Oh, this actually doesn’t work for me. The demo was great. It looks really good when you show me the demo on the iPad. But when I’m at [00:07:00] my, my Windows XP machine in the back, it doesn’t look like that.

Erik Sunset: It doesn’t work like that. So I’m going to jump ship and move on to the next one. You’ve got practices now that are on their third and probably final EHR. Cause to your point, why would I change? I’ve already seen three of them. None of them are going to make me happy. So I’m just going to stick with this one.

John Williamson: And it’s hard. They do not. There’s there’s strategic lock in happening where if you make it difficult for you to get your data out, You will not be able to go somewhere else, even if you want to, and everyone, all the vendors know it, uh, and yeah, it’s sad, you know, it really is the, it’s unfortunately one of the, it’s sort of the cornerstone of the sort of the sequence of events that has to happen for sort of healthcare to work.

John Williamson: The EHR is really a cornerstone of that, and it’s, it’s a terrible one, and that, that’s unfortunate.

Erik Sunset: I have one more thought on the topic and then we can move on that. There’s even a pretty recent law around data blocking. So that, uh, an EHR [00:08:00] vendor can’t hold data hostage. There’s other provisions to you. This wasn’t pointed straight at the EHR vendors. But even with this law in place. You know, you’re taking, in a best case scenario, if you can get it, discrete data in your old EHR, so data that you can report on, data that lives in a particular field, like think about vitals fields, you know, height, weight, blood pressure.

Erik Sunset: If you can get it, you’re moving into a PDFed version of that same chart note, which does you no good to report on anything. So you even have laws in place that can’t fix this issue. And that’s a whole other topic. Do we lean on the federal government to write laws to fix our problems? I don’t know. I think the vendors can do better, but that maybe that’s neither here nor there.

John Williamson: Well, I think, I think there’s going to be, you know, I, I tell the story of this sort of black phone, black grade iPhone. You know, switch over and where everybody thought, well, of course, Blackberry is going to win forever. The physical keyboard, everything what’s this, you know, goofy nonsense, you know, who, who needs a web browser on their phone to then [00:09:00] there is no Blackberry.

John Williamson: And I think if it, by the end of the decade, if this doesn’t happen, I will be wildly surprised, but that the long tail. Many of these are going to switch not to an existing EHR vendor. I think there’s already this work happening. I’m sure there’s, you know, well qualified startups working on this, too. Sort of an AI first EHR experience that abstracts away a lot of the button clicking in the field filling because most of that sort of traditional database, you know, sort of a U.

John Williamson: I. On top of a database that that really isn’t what health care needs. It’s what Tools were available, but really, you know, healthcare is all about context and, you know, the context of a person. It’s not just the, you know, the soap note for this visit that matters. It’s also what has been happening over the last three months.

John Williamson: Maybe even some of the, you know, what have they been, you know, buying or not buying, filling prescriptions at Walgreens and those kinds of things. The context is what matters and sort of an [00:10:00] AI first approach. you know, they call it the context window, right? It’s the context window that you want. So as a, as a patient, what you really want is the ability to sort of Take everything that this doctor’s learned about me, plus everything I can tell it because I know my health, you know, in a lot of ways that my doctor doesn’t.

John Williamson: And maybe I have another doctor I’ve already seen. I want to take everything I know, plus whatever I’ve been handed, put it all together. And that’s context. And I think that a lot of this sort of nuance of what we do and the button clicking and the eligibility verifications and all this goofy stuff that we do.

John Williamson: And you know, our whole company is built on facilitating. I actually don’t think That’s going to hardly be talked about when we have that next iPhone moment. I think a net new EHR will be the replacement vehicle, and the fact that everything gets spit out in PDFs won’t matter

Erik Sunset: Yeah.

John Williamson: anymore.

Erik Sunset: Well, that, that brings me back to an anecdote with, uh, some cardiology clients, you know, probably 10 [00:11:00] years ago now. Amazing to think that, but this was when personal fitness trackers were getting really popular. So it was the Fitbit then. And now you have even more advanced tracking on your Apple watch or with the whoop device, but going back 10 years in time, a cardiologist would have a patient wearing a Fitbit patient and say, Hey doc, can you do anything with this data?

Erik Sunset: You know, I’m wearing it every second of the day. And the answer was absolutely not. Number one, I don’t have a way to map this data into my EHR. And number two, think about the tremendous liability I now assume if you give me essentially a continuous EKG for the duration you’ve worn this device, and I don’t find a problem that you had, it’s going to be my fault.

Erik Sunset: But with, with AI, that context window, now you can process huge amounts of data, um, shield yourself from liability to a degree that, look, we had the machine crunch these numbers against these lab results, against my entire patient history with this patient, and maybe you can get a meaningful, uh, insight out of it.

Erik Sunset: Whereas 10 years ago, there’d be no way. Nobody would even touch the data.[00:12:00]

John Williamson: You know, even the data that does get put in there, how often have you been in a healthcare experience? I’ve seen this when I’ve had to interact, especially with, you know, babies being born and these kinds of things where this person will come in And just look at the chart basically for the first time while they’re standing there and then ask you to tell them the story of what else has been going on that in theory is also in the EHR, you know, just, you know, a click or 2 away from where they are at the moment, but I think we’ve actually gotten into this, um, you know, tunnel vision.

John Williamson: It’s almost like we’re, you know, we’ve got to. You know, an encyclopedia of a patient and their problems and sort of every everything the health care and medical industry knows about this, and we pull out a telescope and look at, you know, one or two words at a time around the page. So, I think even the usage of maybe over reliance on the EHR and because the way it’s set up splits the data up so much [00:13:00] that I don’t actually think providers are as well informed today as they used to be.

John Williamson: Because they could at least see the better, have a better historical context, even if it was just paper, and I’m not advocating for going all the way back to paper, but I think where we are now, unfortunately, we had the worst of all worlds. It’s, you know, tedious, expensive, and it’s also not helping. Us You know, the, the reason that the, um, you know, the, there’s this whole, uh, thing going about risk score inflation, and there’s a lot of incentive for these Medicare Advantage plans to increase the risk score of their patients by applying sort of unreal, you know, undocumented diagnosis codes that the risk weighted ones and providers and value-based care are doing the same thing.

John Williamson: ’cause risk management is the easiest lever to pull between quality, cost, and risk. Uh, and the, the reason that is even a thing. Is because of lack of context. So you came in and I didn’t write down that your diabetes, you have diabetes, cause that’s not what you’re here for. But somebody’s incentive is to say, Oh, well, [00:14:00] contextually they have diabetes and therefore their risk score is higher.

John Williamson: So even just the process of having the EHR, we’ve lost, we’ve lost context of what’s going on. What is the holistic patient? And I think the EHR is, you know, a good chunk of it, of the reason.

Erik Sunset: I believe it. And I will be the first to, uh, share fault where fault is due with the, with the vendors, with the way the UIs are laid out with the entire user experience. And what you’re hitting on now is I know the core of our episode today is around the intersection of AI and automation into these legacy EHRs.

Erik Sunset: Like you said, the vendors aren’t going to throw you this party. You need to throw it for yourself. But it really extends beyond just EHR, like we’ve been discussing, to the clearinghouse and then to insurance portals as well. So what are your thoughts there?

John Williamson: Well, uh, you know, I, I’ve seen this. Play out often. Uh, I used to work at a clearinghouse and we had an analytics product [00:15:00] and we used to kind of internally say, well, why don’t more people use these analytics? And then you talk to a provider and what the source of truth is in my EHR. But then you’re in your EHR and in the real world of their well, Yeah, it’s kind of the source of truth, except it’s not up to date because, you know, you don’t post zero pay remits or, you know, you didn’t capture the recoupment or you have no idea what’s going on with the adjudication of a claim that stuck and maybe someone made a phone call and left a bad note or didn’t leave a note or whatever.

John Williamson: All these things aren’t, none of it individually is the source of truth. So the reality is, you have to sort of pull together multiple sources of information. aggregate them together, make some sense of it, and then you can evaluate individually. None of those data sources actually tells you the full story, even of an individual claim.

John Williamson: Forget the patient, their health and outcomes, but just a claim getting paid. You still can’t do that with just [00:16:00] one of those systems. So you have to be able to work across them. Obviously, this is what the individual you know, rev cycle workers are doing all day long. It’s so inefficient and there are gaps in the process.

John Williamson: So, um, you know, there’s a number, for example, of insurance companies where they either a don’t have a portal or the portal they have doesn’t include all the relevant information for when there’s an exception. So they invite you to call. They don’t want you to call. They won’t answer the phone for 45 minutes.

John Williamson: But that’s the only way to get the information that’s going on. So they’ve created an environment where everybody’s running around so much to sort of get what is the current state of the union across all the areas that data resides.

Erik Sunset: Well, and this is through a little bit different lens, obviously, the payers are using AI to respond to claims. I know how I personally feel about the AI denial for no good reason other than that we’re going to deny claims [00:17:00] the first time we see them and probably the second time, maybe the third time too, but it seems to me that the practice or that the healthcare organization really needs to Increase their own armament as it were with AI so that they have a whole being able to unify data and then being able to respond to hiccups in the claim process without needing to involve FTEs that they either don’t have on hand or that we better serve doing something else, something more productive for the practice.

John Williamson: Yeah. Yeah. And the F. T. Unfortunately, you know, the industry that we’re in, in this sort of administrative side of health care, really in the clinical too, but the administrative side is so big. There are not enough qualified people and the pandemic didn’t help with this. Um, there was huge surges in, in sort of, you know, uh, salaries and costs of doing work.

John Williamson: And, you know, when it. Taco Bell is paying 19 an hour to, you know, stand there and push a button. [00:18:00] Are you really going to be able to hire, you know, an AR rep for 17 an hour that has to learn and do all this, you know, tedious stuff. Uh, and so the answer is no. So that’s sort of jacked up the costs on everything.

John Williamson: Plus there’s not any, there’s not enough qualified people out there. The folks that are, are retiring a lot of work moved offshore, of course. And so we’re in the, we’re in the worst case scenario right now, unfortunately.

Erik Sunset: And that’s an interesting split, at least to me, sort of the pre COVID ethos of just throw FTEs at a problem. We have something where there’s a throughput issue. More people can fix that. You end up with a slightly better outcome. I would argue that was the wrong approach pre COVID, post COVID at any time, because there’s workflows and healthcare that the rest of industry would never accept.

Erik Sunset: And sort of the example that comes to mind in a consumer sense is you want to watch something on Amazon Prime, you open it up on your TV or your iPad or whatever, you click it and you’re watching it. But [00:19:00] the healthcare equivalent would be you click to watch it, and then you have to call a phone number for it to actually play. So now, post COVID though, there’s nobody to hire, as you pointed out. There’s a huge crunch in staffing. Practices and ASCs are having staff gobbled up by the hospital that can pay a little bit more. You certainly lose out on some of the culture, if that matters to you as a, as an employee. Some of you selling your time, uh, you know, in exchange for a paycheck, obviously.

Erik Sunset: Um, But you’re having to do more with less than you’ve ever had, and probably less than you’ll ever have again. As a, as a healthcare organization. So where, where does AI fit into that? Or what would be your words to the wise for an organization that knows they can do more and do better and is ready to embrace a digital outcome where previously they hadn’t even considered it?

John Williamson: Yeah, yeah. Well, and you want, you want to give your staff leverage. So, you know, whether that, that leverage is typically in the form of automation, uh, you know, there’s some automation that sort of takes no [00:20:00] thinking at all. It’s, you know, have a bot go here and download this and have it go put in over here and change this file or whatever.

John Williamson: There’s that, which is useful. And then there’s the other kind, which is okay, now interpret everything and make a suggestion about what I should do next. Uh, you know, both of those things can give leverage to employees. If it’s baked into the sort of operational workflow, and this is where I think one of the biggest challenges that sort of sort of, you know, next level deeper sort of below the surface of the answer.

John Williamson: The iceberg problem is that getting a person on your team to change their workflow to incorporate another tool is a big challenge. And even if you have brought leverage, you know, sort of to the party, if you don’t get people to use it and leverage it, you know, take advantage of that, uh, as much as they can, then you’re not going to get the R.

John Williamson: O. I. That was sold on there. So you have this workflow problem because folks are working across so many different systems and screens. [00:21:00] The real question will become, can you delete a system from their workflow? Can you make it so they never log into the clearing house again? Can you make it so they never go out to a payer portal site again?

John Williamson: You know, all these things, right? Uh, can you make it so even, you know, at the nth degree, sort of what, you know, the vision that we’re chasing down on our side is, I automate so much of it that they essentially make a decision and don’t even touch the EHR? Can I, can I, you know, sort of abstract and obfuscate.

John Williamson: the systems by gathering all the data from all the places, presenting it in sort of decision ready format decision is made button pushed and then another bot takes over and does the action. You know, that’s where you got to get to, to ultimately to be able to process as many claims as there are to be processed, you know, with the increase in, to your point about the, you know, adjudicate AI assisted adjudication.

John Williamson: Which, you know, whether it’s they found a technical nuance that they can sort of, [00:22:00] you know, recoup the last two years of payments based on some, you know, NPI, blah, blah, blah, mismatch thing, or whatever, or it’s the, you know, well, our current latest best practice guideline includes this charge codes are usually have these diagnosis combinations and therefore you didn’t jump through the right hoop.

John Williamson: And I can, I have more excuses to deny it. So my my rationale and reasoning for denials are cranking up. So the sort of response time for exception management on the on the provider side. You’ve got to get more leverage, more people or more leverage, sometimes both to be able to handle that. And, you know, like we talked about earlier, it’s not like the tools that are out there already are helping.

Erik Sunset: Uh, no, not at all. And the tools that come to mind, you know, the legacy stack of tools is the response code and the PM. So not even going into the clearinghouse, but we saw this claim had, it was denied. And now I got to go into the clearing house and see [00:23:00] why. And then now I need to go work this denial and then hopefully push this claim through to the end of the process.

Erik Sunset: And we can be, we can be reimbursed. I mean, it’s, it’s an arms race is what it is. And, you know, you certainly need to do more with less staffing at the health care organization level, but you can’t do it with no staffing. And you certainly need to have sort of this culture of change, not just embraced from the top, but driven from the top so that there’s an understanding and a culture.

Erik Sunset: It’s kind of funny, we’re talking about AI and software technology and digital health, but none of that matters if there isn’t a culture of using these tools at the organization, right?

John Williamson: Yeah. You know that we were kind of using the phrase bringing a party, right? Because, uh, we, we’ve learned this lesson ourselves of there’s a, uh, a friend of mine that, uh, has this great sort of joke. He’s told me and I, I like to repeat it. You know, you could. Park a spaceship in a field [00:24:00] in Iowa and leave the keys in it and walk away and come back a month later.

John Williamson: It’s still going to be there. It doesn’t matter how amazing the technology you bring is. If no one uses it, it doesn’t matter and it won’t use itself. So you really got to Bring the party, you know, including like, you’re going to come and dance. Like you’ve got to really be involved. So, you know, I’m thinking that there’s this, um, phrase that’s becoming more popular.

John Williamson: You’re very familiar with sort of software as a service, the sort of updated phrase is service as software. And what that really means is that you’ve got a, uh, a tech enablement with a human in the loop component. And you bring the human meaning you bring the party. So instead of trying to get. You know, so and so person who’s been doing their job the same way for 20 years to now fundamentally change how they work so that they can take advantage of the leverage you brought.

John Williamson: You’re going to need to do that for them, and then they just get the final product. So instead of, [00:25:00] I think what you’re going to see a lot more of in the vendor space is where they bring a tool and a super user of the tool, the human in a loop component, because sometimes it’s pretty technical. Sometimes, you know, the up, the training to sort of do all the things that need to do is too high.

John Williamson: And so you’re really just going to go solve the problem, not sell a tool. And I think that the sort of mindset of, you know, the CIO CTO world is also looking for just because there is new technology that I’m going to have to support and adapt and train and blah, blah, blah, work, work, work. Why don’t you just do the thing?

John Williamson: Why don’t you just solve the problem? Don’t just bring me a tool to help me solve the problem because then I just have to do more work to do it. And you’re better at it anyway because you’re all, you know, your number one focus is this one thing. And so you’re gonna be more efficient at it than my people will ever be.

John Williamson: And then you have sort of guaranteed adoption. Because I don’t think you can automate half a job. [00:26:00] This is this is actually a fundamental flaw. I’ve seen this a bunch of times where someone will go decide that they’re going to automate one thing and they’ll use one of the several tools that are off the shelf tools.

John Williamson: They’ll hire a consultant to help them. They’ll they’ll build it and then it doesn’t get used on a consistent basis because it assumed that there was a sort of qualified driver of that automation. And integrator of the automation into the, you know, sort of the final process and so often there’s not.

John Williamson: And so you have a lot of sort of false start failed automation projects and one offs that people get burned and lose their job over often, where a lot of nothing happens. And unfortunately, that’s happened so many times, maybe in the last 67 years since sort of automation took off and sort of health care that People don’t buy the pitch.

John Williamson: The pitch is more complicated. No one believes that what you sold them is actually what they’re going to receive because they know they’re not as good at using it as you were at demoing it.[00:27:00]

Erik Sunset: Wow, that, that opened up some new pathways in my head. I have never heard it framed that way, but that makes. Perfect sense and in this future, and I want, I want to hear your thoughts on how far away this future is, but because of all of the efficiency gains that this one expert user would bring to a practice or a hospital department, or you pick your place of care, do you envision a future where there’s one person that floats around a handful or two handfuls worth of healthcare organizations plying their trade as it were in this one specific problem?

Erik Sunset: And then you’re able to keep staffing very light. Um, across these organizations, or would this be more dedication in a full time capacity for one organization?

John Williamson: You probably have both. I mean, depending on, but I think what you’re going to see is a piece of software that, you know, has value all by itself, but also supplemented by, you know, a fraction or more than one in some cases, humans behind the system that sort of [00:28:00] coordinate, enrich, fill in gaps, provide human in the loop stuff.

John Williamson: Because a lot of automation, just like, you know, not every AI Response. If you guys are, you know, using the tools out there, you know, that’s okay, but not really right. You got to have someone look at it and it’s not right all the time. The same thing happens in automation land. Well, you know, a pop up came up and so this thing didn’t flow like somebody’s got to go address it almost in real time.

John Williamson: And you can’t expect an organization to become a super user of your tool. You almost, I really do believe you’re going to see software sold with services behind it and invisible. So there won’t be a physical human sort of onsite, you know, looking over somebody’s shoulder, you know, did you check the clinical documentation to make sure it included this blah, blah, blah, no, it’ll just be, you know, invisible, right.

John Williamson: And you’ll get back. finished product. And you’re going to probably see that, you know, chunked up everything from, you know, patient access and authorization and sort of coding documentation [00:29:00] to, you know, billing and so on. You’re going to see this where what I believe are chunks because the existing sort of foundational EHR, let’s assume it’s still there.

John Williamson: Nobody’s changing. Clearing house does very little, nothing’s changing. Portals, they evolve, but generally they’re static too. That, so the, the sort of environment in which we live and work in is going to stay the same. What I think you’re going to start to see are more vendors that bring a, a solution, not a product.

Erik Sunset: That, uh, that brings to mind one of the worst, you know, sales and hustle culture stories out there, you know, around selling like a drill, right? You’re trying to get somebody to buy this new power drill and it goes however many RPMs and it can hold all these different bits and it can be green or it can be red or it can be yellow.

Erik Sunset: The battery is this size. It holds this much charge, but at the end of the day, people want to put holes in stuff and drive screws or whatever. So sell them the hole. Don’t sell them the drill. [00:30:00] That’s a terrible, I cringe every time I see it online, but that’s what we’re talking about here. I don’t need your software.

Erik Sunset: I need you to fix my problem. And if the software does that, I’m a buyer.

John Williamson: Yeah. I mean, can you give me, you know, a perfectly filled in, you know, clinical chart and coded properly. So I didn’t miss the thing. You know, can you give me the final draft? I think that’s what, you know, you kind of think about it in drafts. So just people just want the final draft, you know, tell me how much, you know, is the patient covered and how much do they owe me right now?

John Williamson: Don’t give me a tool for me to go figure that out. Not that that wasn’t a good idea. Okay. To empower people, you know, sell them the tool, but because, you know, sort of, you know, revenue cycle, healthcare, it’s like death by a thousand cuts, you can’t cover that many bases and you’ve got the dwindling staff and blah, blah, blah.

John Williamson: You just want, you don’t want to, some people have outsourced everything and that’s fine, but a lot of things that’s not going to happen and you don’t want to do that, but you need to [00:31:00] sort of augment and replace, you know, components. So you just give me the final draft and put it back in my EHR. I don’t want to look at your tool if you meet with me on a monthly basis and show me a chart.

John Williamson: That’s great, but I don’t ever want to log in myself. I don’t want my staff to have to learn it or log in or talk to me about it, why they can’t log in or ask me a stupid question. I don’t want any of that. I just want the output come. And I think that sort of that that’s really how you’re going to see the implementation happen of some of the more advanced tool set is that you don’t want to go become an expert in how to.

John Williamson: You know, design a rag system on your own to like perfect this one little process. You just want to have final product. And I think that’s the new cohort of vendors and supplied solutions are going to find ways to deliver the final product at the cost that they used to deliver just the software, because they’ll implement all the AI and automation in their stack and [00:32:00] sort of fully optimize it because they’re focused on it and get it to the point where, Oh, now we can do a done for you solution.

John Williamson: Our cost of goods sold very similar to software only with, you know, support and training and all this other stuff. And we can upsell the client to, you know, turn key. And everybody just wants that now. Nobody wants to buy tools. They just want things. So I just want my picture hung. I actually don’t even care if there was a drill or hole or whatever.

John Williamson: I just want the picture of my kid on the wall,

Erik Sunset: That’s really well put. And that, um, that’s kind of bringing to mind some interesting discourse around the more consumer driven AI products or, uh, LLMs that those are very quickly becoming a commodity. And in the news this last week, a lot of people thought the new release of the X product Grok was blowing away Chad GPT and that Claude had great game ground and Somerset respects, but not others.

Erik Sunset: But these are very quickly reaching that [00:33:00] commodity level because they’re all fairly similar. The outputs are all. Just about the same, you know, blank output into a Word document. Would you be able to tell which one was Grok, which was GPT, which was Claude? I couldn’t, so I’m sure somebody could. Well, what people really want is the outcome.

Erik Sunset: And then secondary to that is like the wrapper, which gets them that outcome. And I think that’s what you’re talking about. Uh, at least a little bit. I don’t want to put words in your mouth, but the wrapper to supply the solution. Is there a human in the loop in the back end? It doesn’t matter to me because I only care about the finished product, the outcome,

John Williamson: right?

Erik Sunset: and it’ll be interesting to see how AI shapes up because, and I’m rambling a little bit here, the other, the other point I wanted to make was that 10 years ago with any type of automation or digital health product, Oh, no, does that mean somebody is going to lose their job at my practice?

Erik Sunset: And even today, that answer is not necessarily yes. But, because you’re understaffed, because you can’t make a hire, because retaining staff is such a brutal endeavor these days, maybe that’s [00:34:00] becoming more and more of a good thing.

John Williamson: Yeah, you know, no one has the problem in my, in my view of gosh, I’m just gonna have excess people sitting around without enough stuff to do. Nobody has that problem anymore. You know, they’re already understaffed and can’t staff. So, you know, in some ways, uh, that. Opens the door for sort of, you know, having a conversation about staffing levels and right sizing because it’s no longer of you’re going to fire people and say, you don’t have to hire.

John Williamson: It’s a different, um, there’s a term that’s gotten pretty popular in sort of the startup culture of static team size and that all of the sort of, you know, more frontier companies are not adding staff because they made their, you know, they gave a tool like cursor to their engineers and now they’re, Okay.

John Williamson: Those engineers are doing better and even the mid quality are doing high quality work and the high quality people are just sort of like machines. And this, for the same reason, that sort of static team size [00:35:00] concept, uh, you know, will apply here as well over time. You know, once you can automate more of these things for them.

Erik Sunset: And on the other end of that, the willingness to embrace these new automations, softwares, humans in the loop or not, it seems like there’s a willingness now in 2025 to embrace technology as a healthcare organization, at least that I’ve never seen before. People are starving for it.

John Williamson: Yeah. Well, and some of these, you know, there’ve been these intractable problems that the tech really wasn’t there to solve of sure you can automate this, but you actually have to read the document. And, you know, your RPA doesn’t do that very well. And so you don’t know, you know, yes, you can go get the document, but you can’t figure out which way to go, you know, left or right after you get it, things of that nature where now you can, or, you know, all these components where it used to be so difficult to sort of get to the final solution, um, that it was more work than it was worth.

John Williamson: A lot of this sort of. You know, [00:36:00] all the, for 10 years, I think it was that I saw it for, you know, all the products were sort of machine learning, predictive analytics, all this different stuff. The problem was the, the false positive was way too high and it frustrated everybody. And then it became too expensive to sort of curate that signal from the noise.

John Williamson: And then it was like, why are we even doing this? And so I think this is where you had a lot of burnout in the world of, you know, buyers buying frontier tech that they, you know, maybe they’re done trying, or they’re just waiting for something that really just takes it out of their hands entirely.

Erik Sunset: Yeah. Uh, hard to argue with all of those points. I’m, I’m right there with you on that, John. And as we, as we kind of get to the end of our, of our time together today, any, any big predictions, um, anything that’s coming down the pike that you want the listeners to hear about?

John Williamson: Well, you know, kind of talking about the, the reality check of these EHRs. And I do think they are going to sort of, there’s going to be a Blackberry to iPhone event. On the [00:37:00] horizon leading into that. I think what you’re going to see a lot of we use this analogy often. Um, a lot of people have TVs at home.

John Williamson: Some are older. Some are newer. They pull up an Apple TV or a Roku box into and that became their new front door. Their TV was just they turn it on and change the volume. And other than that, they use the other remote. The bolt on system became the system, the default system, and I think that’s what you’re going to see.

John Williamson: That’s what we focus on right now at lockbox is sort of being almost like that bolt on system that has more of the data than your systems have and can operate and can leverage automation and sort of predictive AI suggestive components in the system. So that you don’t have to change anything about what you do.

John Williamson: You just use the bolt on that’s been optimized using all the new tools. I think the sort of bolt on intelligence, um, especially one that can sort of pull together the output of multiple other things. Um, [00:38:00] I think you’re going to see a lot of that. I think you’re going to see the Apple TV Roku box, uh, strategy being the right strategy here.

John Williamson: Uh, because and that’s where you’re gonna bolt on all your new things. So you’ve got this thing over here, you know, you got DocBuddy over here doing, you know, transcription and taking clinical content. You’ve got, you know, a third party, you know, coding rules that are sort of really optimized around your payer mix and not forgetting the risk weighted diagnosis code and looking backwards and all this stuff.

John Williamson: And you’ve got RPA going out to payer portals and you’ve got patient statement billing and you’re not going to plug those individual items. into your EHR because it’s too painful. You’re going to plug it into something that’s more of a bolt on, like add a new app on the Roku screen. You’re going to plug it into something that’s designed to take augmented data, normalize it, and then make decisions and push it out where it needs to go.

John Williamson: So I think you’re going to see this sort of like secondary hub. Component come online. That’s what we focus on at lockbox. Um, I think that’s generally speaking what you’re going to see, and then you’re [00:39:00] going to see a switch over to, there will be sort of net new EHR that takes the hub components, that strategy kind of like now, if you go buy a new TV, Roku is built into the TV.

John Williamson: Right. That will be the shift. It’ll go from a bolt on that you prioritize to it is the TV that that’s the change that’s going to happen. I think, um, you know, and starting to happen already. It’s going to continue. And I think in, you know, by the decade end, we’re for sure going to see that

Erik Sunset: Love it. Love it. I’m, uh, I’m very tuned into the space as are, as are you, John, I agree. I agree. I mean, the big prediction from so long ago was specialty specific EHRs. We already kind of touched on that. That didn’t really come to pass. They’re all kind of the same, you know, beneath the surface layer. So these specialized solutions resulting in a hub like EHR that handles all of your specialized solutions.

Erik Sunset: I like that. I could see us going there and for the provider’s sake, I hope we do. These EHRs are killing the profession, which we didn’t even talk about. Burnouts. And, uh, you know, [00:40:00] there’s plenty of information about there about that.

John Williamson: stop typing, start looking at me, you know, it’s that,

Erik Sunset: Yeah, right.

John Williamson: Let’s see. What are you? It’s like when you go to the airport and they’re like, just type in and type in. You’re like, what? I’m here. What are you just checking me in? You’re like, what all are you typing? I kind of want to ask the same to the physician.

John Williamson: Like, seriously, what are you typing? I actually know. And it’s horrible. Actually, you probably don’t want to have to tell everyone. I just like this drop down again that you’re still smoking, whatever, like

Erik Sunset: Yeah. Right.

John Williamson: nonsense.

Erik Sunset: I’m on a drop down eight of 10 to start my conversation with you and then after that I’ve got five more for your physical exam. I mean, it’s just, it’s ugly. It’s ugly.

John Williamson: It’s terrible. It’s terrible. I’m glad solutions like DocBuddy are out there to help with that too, because, you know, that, that through point of patient care where you have that lack of interaction that that’s, you know, that’s one of the main touch points in our healthcare system that’s been negatively affected by the implementation of clunky EHRs.

Erik Sunset: We’re, we’re doing our small part to, uh, reduce symptoms of burnout across the space. Obviously, um, we need to [00:41:00] do a lot, a lot more of it. The industry needs to do a lot more of it, but getting physicians away from their computer is the number one thing that needs to happen to make healthcare better for everybody.

Erik Sunset: There’s a lot to be said around payers and how they treat their providers. Cause that’s, that’s a whole other ball of wax, but at a baseline, we’ve got to take better care of our providers or we’re not going to have any, uh, anybody to treat us by that decades end mark, like you’ve referenced.

John Williamson: Yeah. True story. True story.

Erik Sunset: Well, John, before you go, where can listeners, uh, follow you online? Where, where can they see what you’re working on? What are some links people need to have from you?

John Williamson: Sure. Sure. Uh, so our main website, lockboxai. com is kind of where we find out about the company. We’re on, uh, LinkedIn at lockboxai and I’m easily reachable at john. williamson at lockboxai. com.

Erik Sunset: Cool. We’ll get all those links and that email address into the show notes. John, thank you so much for joining us today.

John Williamson: Yeah. Thanks a lot. Appreciate it, Erik.

Erik Sunset: Anytime you’re welcome back anytime. And on behalf of the entire DocBuddy team, I want to thank [00:42:00] you for listening. Be sure you’re subscribed on Apple pods, Spotify, and YouTube. So you always get the newest episodes of the show.

Erik Sunset: And until next time, I’m your host, Erik. We’ll talk to you soon.

John Williamson: Thanks.