Latest / Digital Leaders Club By Ingeno / Jeff Cripe, CEO at Procode AI | Healthcare AI, Medical Billing & Revenue Cycle Innovation
Transcript
- Ingeno: Welcome to Digital Leaders Club. I'm Guillaume Duchenot, founder and CEO at Ingeno. The show is built for the leaders shipping real world products into hard markets. And today that means someone building technology for the most complex industry there is. I'm joined by Jeff Kripe, Chief Executive Officer at Procode AI. Jeff, welcome to the show. Jeff Cripe: Thank you. Nice to be here. Ingeno: All right, Jeff. so before we get into your company and and operational stuff, tell me how did you end up in building healthcare solutions? Because also almost nobody chooses this market casually, right? You kind of you're brought to it somehow. Jeff Cripe: Yeah, I actually agree with that. yeah, so this is the first company I started eleven years ago was in the commerce and and advertising space, also venture backed, so completely different from healthcare. and you're right, most of the people I've met in healthcare have a personal or a familial relationship to healthcare in the sense that somebody was probably afflicted by some devastating disease or a traumatic accident or their family is, you know, sort of been a long line of physicians and they've chosen to go into healthcare in some way as a result of that. Mine was a little more serendipitous. I bought a house in twenty twenty one, and the only other kind of young guy my age in my neighborhood, and this was during the pandemic. so yeah, we were kinda looking for friends and trying to socialize and you know. a normal way, but it was kind of hard to meet friends otherwise. And we just moved back to LA. Ingeno: Mm-hmm. Jeff Cripe: but anyhow, I I I I met my neighbor who's now my co-founder and one of my closest friends, he's a plastic surgeon in a in a private practice based in Beverly Hills that does both insurance and cosmetic work. and as we were having dinner and the drinks, you know, over a couple of years you know we both talk about work a lot and one of the conversation topics that kept coming up was how frustrated he was with his various medical billing providers. And he'd been in, you know, in six years of practice, I think he'd been through three firms. And he was sharing Ingeno: Mm. Jeff Cripe: with me the different pain points he was experiencing. And it just seemed like pretty obvious problems that could be solved with technology. And Ingeno: Mm-hmm. Jeff Cripe: the more that we looked at the space, we realized that the market was really Both big and interesting and ripe to be disrupted. There's, you know, revenue cycle management is probably a $50 billion a year industry. it's super barbelled. There are big, obvious companies at the top like Ensemble and R1 and AGS. There's really not much of a middle market between call it 35 and 200 million of revenue. And there's a massive Ingeno: Mm-hmm. Jeff Cripe: but heavily fragmented long tail of small RCM companies in the US. And There's a ton of innovation going on to support the big guys and the big healthcare systems. And there's historically been very little innovation going into this long tail to support smaller RCM companies and the smaller practices that they service. And in addition to that, those companies really struggle to scale. That's why there's no middle market. 85% of those 5,000 RCM companies generate less than $5 million per year of revenue. So there's a ton of companies. They need to be consolidated in some way. It's hard to structure peer-to-peer transactions and there are structural issues that are preventing those companies from scaling into into middle market companies. And we basically looked Ingeno: Mm. Jeff Cripe: at the industry and talked to Cam and understood his pain points and basically said, look, if I can surround myself with healthcare talent, I've got a vision for what the product is and how it can drive automation and margin enhancement within these businesses that we would acquire. We can both be that technology and be the liquidity that consolidates that space. so it was, Ingeno: Hmm. Jeff Cripe: really started over drinks drinks and a dinner and but it's a it's an industry I've come to come to love and really enjoy working in. Ingeno: Awesome, awesome. so you touched a bit on it already, but can you tell our listeners about Pro Code AI? What you guys do exactly, the product, who's it for, the scale, yeah, where are you at today? Jeff Cripe: Yeah. essentially Pro Code AI is you know, AI roll up has become a term that, you know, I think has broadly been embraced by General Catalyst and Insight and you know, Lightspeed and a number of other a number of of the other kind of leading venture firms. But you know, when we started the the business three years ago, there was really no concept of an AI roll up and it it's kind of like a a strange hybrid of of of a venture strategy combined with a PE strategy. So we looked at the market and said, hey, we think it's gonna take a long time to sell these products that we wanna build into RCM companies. We think there's gonna be low average average contract values, and we think it's gonna be hard to make that sale and integrate our product because those companies don't have technology teams. So if we want to penetrate that market, the only real way we thought we could successfully penetrate it is by acquiring the companies themselves and vertically integrating Ingeno: Mm. Jeff Cripe: that technology. And the benefit benefit of that is instead of selling a per seat, task contract, we actually get to r capture the full value of the margin improvement that we can drive. So we've got three core components of our technology suite. one is an AI coding engine that automatically codes operative reports, you know, all the charge codes, billing codes, diagnostic codes, modifiers, units, etc. And we do that with a very high degree of accuracy. that both maximizes reimbursement and is is obviously compliant and speeds up our team and increases our productivity and drives faster payments for our clients. we've built an operating system for our billing team which gets them out of spreadsheets, centralizes them all in one workflow software that is built on top of a unified and structured database pulling from all the EHRs that our clients work with. and the third thing is a prov a provider app. Historically the way that RCM companies communicate with their practices are through spreadsheets that sometimes they update and sometimes they don't. We put an app in your pocket, think of it as like QuickBooks for Surgeons that shows you all of your revenue trends, your payment trends, your charges, payments, and adjustments, how those all compare to you know year over year, month over month periods. We show you revenue by payer, we show you revenue by CPT code. We show you all of your AR aging charts by by volume and and and by outstanding revenue. We'll show you d you know DRO, we'll show you top unpaid cases, and we have a bunch of other tools built into that app as well that surgeons frequently use to make it easier to translate or transmit data to us and you know generally just sort of go about their day. So those are the three core pieces of technology and we're basically acquiring these companies. And then in and then integrating those p pieces of technology, those components of our platform into the companies that we acquire to drive better results for our clients and better results for for us, obviously. Ingeno: Nice. So Jeff, can you explain what a traditional RCM company does and how with AI and technology you make them be in the 21st century, let's say. Jeff Cripe: Yeah, the the simplest way to describe it, and you know, as a non healthcare person, I've had to figure out ways to explain these things to myself because I didn't know what an RCM company was four years ago either. But the simplest way to think about this is let's say you're a surgeon, you're an orthopedic surgeon and you're I'm your patient and you I get into a skiing accident and you know, I call you and I say, Yom, like I you know, I I just I think I I tore something in my knee and I come in and I see you and you say, Yep, you tore your ACL, you need to get knee surgery. And I schedule an appointment with you, and you do that that surgery on me, and I leave, and it's covered by insurance. Now you need to get paid for that. but for the most part, surgeons don't do their own billing. So they rely on third-party billing companies or our revenue cycle management RCM companies like ours that basically sit in between the provider and the payer, the insurance company. and advocate on their behalf to get them paid. so Ingeno: Mm-hmm. Jeff Cripe: they handle all of the steps from coding that operative report to scrubbing that claim, submitting that to the clearinghouse, onto the insurance company, dealing with denials and rejections as they happen, in our case also handling patient payments, posting payments once they come through. all of those administrative steps in between you doing a surgery and you getting paid for that surgery, it's just a cycle. And we handle all of those steps along that cycle. that's the easiest way to think about it. Now it's a very labor-intensive business and it's actually like a highly skilled business. So if you're in a coder, right, you're the person that historically is a human who has to read that operative report that a surgeon produces at the end of that surgery. There's a lot of very technical medical language in that document that, Ingeno: Mm. Jeff Cripe: if you are trained properly, indicates to you what alphanumeric codes you should use to build the insurance company. And then the insurance company is responding to those codes and will pay based on those codes. So typically people are good at coding for one or maybe two or three specialties, but it's very Ingeno: Mm-hmm. Jeff Cripe: nuanced. Most people can't code for multiple specialties, but most billing companies support multiple specialties. So you run into Ingeno: Right. Jeff Cripe: this problem where as you expand and you're a billing company and you don't have much of a go to market motion, you're relying on referrals. So maybe I started with orthopedics, but suddenly I get a reference for, you know, a referral for an OB, you know, and then a pediatrics, then mental and behavioral health, and then a plastic surgeon. I want to grow my business, so I take them on, but now I have to hire another coder because they don't know how to Ingeno: Mm-hmm. Jeff Cripe: code for plastic surgery. Or they don't know how to work in that EMR that's specific to plastic surgery, or they don't know how to bill and adjudicate some of those claims. So my team gets very big very quickly. And then I say, whoa, it's so expensive, my margin is decreasing, Ingeno: Mm-hmm. Jeff Cripe: I've hired all these people. I now need to offshore this business. I need to bring my labor costs down. So you bring your labor costs down, but then the quality of your work, not in all cases, there is terrific offshore labor, but if you don't source that labor appropriately, which is difficult to do, the quality of your service goes down and you get churn. So in a nutshell, what we are solving is hey, it takes too many people to produce a dollar of revenue in this space. And we are building technology that makes people more efficient, not so that we can replace people, so that we can take a team of a hundred people instead of being able to produce three million dollars, we can produce ten million dollars. That's the goal Ingeno: Mm-hmm. Jeff Cripe: is is to turn these people into superhuman at what they do. So in the past year after acquiring our first company, we grew revenue two X, we grew EBITDA margin five X, and we've grown the team only twenty percent. So we've grown the team, we haven't laid anybody off, but we're getting a lot of more operating leverage out of that team through technology and through management and operational adjustments that we've made to accommodate that technology. Ingeno: That's that's that's awesome. That's that's great. that's really the power of technology augmenting the humans. I I love it. so Jeff tell me what does a typical week look like for you personally? How do you spend your time, you? Jeff Cripe: Yeah, look, I think every CEO has superpowers and it's up to them to sort of figure out what theirs are. mine is all the different ways in which selling is applicable to a business. So raising capital, recruiting talent, you know, are two major ways. And then, you know, building a product vision and selling that into my team and getting their operational buy-in. That's like most of what I'd spend my time on, I would say, is like Raising money, recruiting awesome team members, and helping us execute on our product vision. what is my normal day look like? I'd say I'm usually up. I have a dog that wakes me up at five twenty in the morning every day now. She's a little bit older and she just has to go out early and earlier. So I'm usually up by about five or five twenty. I work until my kids wake up around seven. I hang with them from like seven to seven forty-five. I'm in the office by eight. I'm on calls probably from eight till three on average and then I've got about two hours to do work. I head home, I have dinner with my kids, I put them down to bed and then I usually work until sometime between ten and midnight and then do the same thing again the next day. so it's you know, call heavy in the afternoon, Ingeno: Prince and repeat, yeah. Jeff Cripe: a little bit of work in the afternoon and work in the evening and repeat it the next day. Ingeno: All right. and what would you say is the part of your product or your your organization you're the most proud of right now? Something maybe the team has built, something you guys have figured out that is really working well right now. Jeff Cripe: Yeah, the I I like to say that our AI coding product is both commercially interesting but also scientifically interesting. you know, so we there were other AI coding products in the market that that are terrific businesses. they've largely focused on E and They've been around since twenty eighteen and they largely focused on E and coding and radiology coding. Nobody really touched surgical coding because it is so technically complex. And those companies Ingeno: Mm. Jeff Cripe: that were started in 2018, 2019, 2020, like the models back then just like were not capable of handling surgical reports even with a lot of training. And we obviously were started in 2024 when you know post-Chat GPT 3.5, you know, benefiting from obviously sort of Claude and its you know incredible advancements. And we basically said, hey, we think we can tackle surgery, and we think this is a really defensible space. So we focused on surgical coding. We're really the the the first and and best AI coding engine specifically for surgery. And we've demonstrated in a in our first research paper that was published in the PRS Global Open, the the primary plastic surgery journal, plastic and reconstructive surgery journal about a month and a half ago, the results comparing the performance of our engine compared to out of the box LLMs, compared to professional human coders. And the results were pretty astounding. We were significantly more accurate, obviously, than out-of-the-box LLMs, and significantly more accurate than human, you know, professional human coders as well. And I think it really just like drove home like, you know, there's I think a hundred and ten thousand ICD ten codes, there's like seventy thousand CPT codes. You combine that with different modifiers and and different units that are required to bill appropriately. There's different payer policies, you know, payers in one state. or a payer might something coded one way in one state and a different way in another, and different payers, differ in how they like things coded and providers also have coding preferences. This is, you know, just a human is inherently limited by their brain capacity and all of the, you know, like infinite number of different combinations of those codes that could occur. Ingeno: Mm-hmm. Jeff Cripe: but AI is not limited by that. And it it is like completely Ingeno: Yeah. Jeff Cripe: academic that like AI should do coding and should do it better than humans. And we should free up humans' time to do other things that AI can't do that are more valuable to our organization. speci especially AR and denials management. So I to me, I think that that's one of the things that's working the best and it's it's it's it's scientifically interesting and and meaningful, but it's also like commercially like an effective product for us. Ingeno: Nice. and did so you mentioned you're using models like Cloud or Ch ChatGPT, but have you developed your own models? Have you kind of fine-tuned those or it's more like a sophisticated way of using existing you know frontier models, fund foundational models, and just organizing their workflow based on each specialty? Jeff Cripe: Yeah, it's look, it's changed over the past two years in particular. probably over twenty-four months, we've really changed our stack meaningfully three times. we started out doing it one way, which I would say was like more highly deterministic. We've now morphed into something that I would say is more reasoning based. so there's there's been a real shift over time in terms of like how we execute. It's a blend of Frontier models and open source models, optimizing for both performance and cost. and it is heavily predicated on proprietary training data that we get Ingeno: Mm. Jeff Cripe: i you in ways that are not available to the general public. You can go and buy de-identified claims data all day. You cannot buy claims data paired with its corresponding clinical data that are both de-identified. That it isn't out there. So if you want to build what we are building, you have to go find that and you have to find it at meaningful scale across all the specialties that you care about and want to code for. And you can get to a certain level with just out of the box, but you can see in our paper, it's not that good. It's probably a little better now than it was, ten months ago when we drafted that paper based on our research. but it's still not good enough to be commercial quality. Ingeno: Mm-hmm. Jeff Cripe: And So I I would say like that y a co the orchestration is is one part of it and the proprietary training data is another that, is pretty unique to our business. Ingeno: Makes sense. And I guess y you get that data because you acquired organizations that had that data, right? work that was coded by humans and that you you know worked, right? Insurance companies accepted those and yeah, makes sense. Jeff Cripe: Exactly Ingeno: then on the the other side, what would be say the the hardest part of your Your reality, your software, your your organization Jeff Cripe: Yeah. Ingeno: right now that someone outside your industry would not fully appreciate. Jeff Cripe: Yeah, it's easy. It's the it's the EHR and PM systems. EHR is an electronic health record system. PM is practice management system. it's great that they exist, which was really a result, you know, of of you know Obama-era legislation forcing us to bring Ingeno: Mm-hmm. Jeff Cripe: all this like health data online. so it's great that they exist. but They are a lot of them are old and they are not designed to make it easy to share data. And the ones that are new, could share data easily, but choose not to and actively discourage that. They're like definitionally like wall walled gardens. Ingeno: Mm. Jeff Cripe: and that makes it really difficult to build in this space. In addition to that, even if you could get access to those systems, their taxonomy, Yeah, taxonomies are completely different. They're they're their their language is completely different. It makes it difficult to structure data that's coming out of different systems. And we sit across, we're pretty unique. We for a company of our size, we we work across sixteen different EHR and PM systems. I'd say most companies our size maybe handle three to five because of the complexity of working with and within them. but we sit across a bunch and it makes building products a that ta that are able to support all of our clients across all of those systems, really challenging and just full of edge cases. Ingeno: No. Jeff Cripe: there was a a one of the founders of Candid told me before we started this company not to start this company. And he said, Ingeno: Yeah. Jeff Cripe: and he said, look what you're gonna f you know, you haven't been in R CM, what you're gonna find is, you know, tackling automation in R CM is like, you know, every pro every issue is like solving a quadratic equation. there's just problem after problem after problem after problem. And he was right. it's it's a huge pain in the butt. But I also think that's part of the mo, right? If this was easy, everybody would do it and if we can figure it out, I I Ingeno: Mm-hmm. Jeff Cripe: think it's gonna be quite difficult for other people to. Ingeno: Nice, I love it. tell me, so you're kind of in between, let's say, the the professional and the the payers, right? How do you how do you optimize, let's say, or minimize denials and how do you can you commit to some to some extent to having those as low as possible for the professional? Jeff Cripe: Yeah, it's it's a handful of things. One is better organization on our end. you know, I'd say than most companies. Most companies are relying on spreadsheets. they don't have a lot of the processes in place, they don't have a lot of the automation in place that we do. If you don't get a claim in on time, it can get denied for timely filing. So there's just like a a process and an automation and a client relationship management standpoint that helps us just one, do the first and most obvious thing, which is not have timely filing issues. And a lot of RCM companies do have timely filing issues. And sometimes that's on the provider because they, you know, you've got ninety days to file and they sent you the, you know, the the the claim on day eighty nine. And sometimes it's on RCM companies because they got it on day one and they forgot to submit it by day ninety. But we have solutions in place that tackle that. The other is coding. If you code better And better means in ways that capture the maximum revenue, but is compliant Ingeno: Mm-hmm. Jeff Cripe: and also are tailored to different insurance policies and preferences, you can get better results. So our engine is trained to do that and you know, as I said, like it's really hard for humans to do. It's just like so much knowledge to try to hold in your head across too many specialties and too many you know, coding combinations and too many payer policies. It's impossible. It's unreasonable to ask a human to do that. AI is really good at that. Ingeno: Mm. Jeff Cripe: I I you know, I I think those are t two of the two of the major ways. and you know, th I those are on the front end. And obviously just, you know, we're able to to build better denial management practices, you know, through automation as well. you know, being able to you know, quickly look at our entire claims history and diagnose, hey, you know. We've seen this issue X number of times before, you know, i in this situation with this payer in this state, this is the the path that usually leads to the best resolution. You know, hey, would Ingeno: Mm. Jeff Cripe: you like me to take action on that path? so tools like that are are you know Ingeno: Nice. So your system can pro can suggest ways to navigate di depending on on the payer in the area and that's that's awesome. Jeff Cripe: Yeah, to d I would say today it's today the AI coding tool is genuinely an AI product, our what we call ProCollect. I would say it's more of a workflow software today with like Ingeno: Mm-hmm. Jeff Cripe: marginal intelligence and automation. we started with the front end of RCM. Now we're moving into kind of the middle and back end over the next, you know, part of the series A that we raised is is to continue investing in the front end, but also to increase our investment in that sort of like middle section of RCM. So that we can bring the same intelligence and automation to that portion of the cycle as we have to the front end of the cycle. Ingeno: Yes. so tell me Jeff, looking forward, if you could fix one thing about how your software runs or your company runs, something maybe that's been on the list for a long time, something that's keep keeps getting pushed, what would it be? Jeff Cripe: Well, yeah, it's interesting. one of the things that I've been thinking about and we're now starting to do some work on is just hey, you know, if we view MA, which we do, as, you know, fifty percent of the way that we're gonna grow this business, and then the other fifty percent would be driving organic growth on top of each of the acquired businesses. Ingeno: Mm-hmm. Jeff Cripe: But if fifty percent of our growth is gonna come from MA, how can we build out an automated pipeline that helps us do MA better and faster? Could that look like know MA SDRs that are constantly prospecting and doing outreach for us and qualifying leads, as we do with our, you know, in our normal go-to-market sales function when Ingeno: Mm. Jeff Cripe: we're trying to acquire new practices? could that automate some of the pre-LOI diligence process? could then, you know, could that then funnel companies to me appropriately to meet with and spend time with? And if we want to get to a deal, you know, work with Harvey or Lagora to actually automate those transactional documents and speed up that deal process. Look, it's aspirational. The way that we're doing it today is still very relationship driven. and Ingeno: Mm-hmm. Jeff Cripe: you know, I would say with sort of like point solution automation where it makes sense, but you know, one of the things we just haven't gotten to, because we're focused on obviously like the core products supporting the RCM business, are, you know, for me specifically, is like how can we start to automate MA a little bit more? now we're not doing such a high volume of that that that's you know, we're trying to do a couple deals a year. So it's not like we're trying to do two deals a month. you know, we'll probably acquire eight to ten companies and you know, from there we've we've got a pretty good base to build organically on. but I do think it would be interesting if we were able to do that, but it just hasn't hasn't risen to the top of our our engineering teams list for you know some obvious reasons. Ingeno: an acquisition engine. So yeah, interesting. Interesting. What would you say is the because you've been working with systems like EHR and and the others that you mentioned, what would you say in the the health tech industry is the biggest opportunity where AI technology can be leveraged to yield better outcomes? So clearly you've you've found one niche, right, where there's there was an opportunity from doing somewhat manual work and automating that. What would you say is another big area where there's gonna be some more than than evolution, like a revolution using AI and data. Jeff Cripe: That's a great question. Look, in our space, I would love if somebody could figure out a mediation layer that can like sit between all these electronic health record systems and make it easier to access structured data. I I don't know what that looks like, but probably somebody will figure that out one day. yeah, generally speaking, like I I don't know. There's just like so many you know, the the the biosciences and you know therapies and and vaccine component of AI is is always like hugely interesting. the bioterrorism like flip side of that coin is also terrifying. Ingeno: Mm-hmm. Jeff Cripe: what else? obviously like, you know, I I was listening to I think in February, I was listening to Vinod Kosla on stage at a conference I was at and you know he was saying in so many parts of the world just access and this is obvious, but living in the US, you just don't see and feel and think about this every day. But in so many other parts of the world, just access to basic healthcare is just not not available, Ingeno: Mm-hmm. Jeff Cripe: right? a lot of people go through their entire lives without seeing a doctor, you know, in in kind of low income countries. And the idea that, you know, in the not too distant future, it it it would be possible to have, you know, AI physicians available to people, Ingeno: Mm. Jeff Cripe: is is a pretty compelling idea to me. You know, a physician in a pocket that like, look, is it the same as as seeing a human? Maybe not, but is it an improvement upon that person's life and kind of current situation, Ingeno: Yeah, yeah. Jeff Cripe: like h hopefully. I like that idea a lot. I mean I just find it like sort of both commercially interesting and like meaningful and interesting on a human level. but I don't know. It's a the healthcare industry is I I think one of the hey healthcare and legal obviously are are a couple of the areas that are sort of like most quickly adopting and you know coding Ingeno: Mm. Jeff Cripe: goes without saying like software engineering goes without saying. But I there's a reason why and it's just it's applicable in so many different areas. Ingeno: Right. Give me, let's say, one technical decision that you that you that you you made let's say last couple of years that you would call a real win. Jeff Cripe: That would that I would call a win. Ingeno: Yeah. Jeff Cripe: Technical. Operational was definitely the choosing the path of an MA strategy versus selling SaaS tools into RCM companies. Like that, that we've we've borne that out. That that would not have worked if we tried to do it the other way. And what we're doing now is is working well. technical. man, hiring my co founder. I worked with him for 10 years at my first company, and it it he was the first person I called, not because he was the only engineer I knew. I'd I I know a bunch. but he was just so uniquely suited to solving this particular problem. Ingeno: Right. Jeff Cripe: that he was he was my first call and we just immediately started working working together. I think there are different ways to partner with these EHR solutions and I think we've figured out good ways to do that after false starting a couple of times on you Ingeno: yeah. Jeff Cripe: know on other strategies. yeah. Ingeno: Hmm. And tell me, you you you kind of decided something really early on, it seems, where the way to scale is to acquire those RCM organizations, right? And you saw that if you try that by yourself, it wouldn't work, right? Was it obvious from the start, or you kind of tried the other way and then figured, okay, it's not working. We need to change. We need to pivot. Jeff Cripe: Yeah, we didn't ever try it the other way, but we spent and but it wasn't obvious to begin with. We only figured it out by spending enough time. You know, as I said, my co-founder had been through three different billing companies. We met with two Ingeno: Mm-hmm. Jeff Cripe: out of the three and got to understand and and you know, really know in a detailed way th their their workflows, the different systems they have to access, how data flows through their system, what the life cycle of a claim looks like. And it like that was the moment when we saw underneath the hood that we said, we trying to the only way to get the full value out of the Pro Code product is to be tied to the EHRs in some way. And it's difficult Ingeno: Mm-hmm. Jeff Cripe: to forge those connections. And for a variety of reasons. Sometimes because the surgeon is operating in a hospital and it's the hospital system. Sometimes because it's you know, the surgeon is actually the client. So to get access to that system, you've got to you've got to go through the surgeon themselves. and and other reasons too. There was just no way for us to scale up the integration of our product with thousands of different RCM companies with Ingeno: Mm-hmm. Jeff Cripe: that didn't have tech teams and all sit on different EHR platforms. And once we realized that, that's when we said, Hey, we actually just have to we have to be the company. We we can't just be the tool. And it ended up being a really good thing. one, you know, the results have been good, but two, as I think about it, you know, there may be a point in in time in the future where we want to commercialize and and sell our product to third parties. we're not Ingeno: Mm-hmm. Jeff Cripe: foreclosing on that possibility. It may come. but let's say we had gone just the SaaS approach. If I'm wrong, and in a couple of and it's possible I am, and in a couple of years Claude can code just as well as we can. not software coding, medical coding, and I'm just a medical coding coding tool, we're dead, right? there's no Ingeno: Yeah. Jeff Cripe: reason for us to exist anymore. But if I actually own the client relationship and I own the outcome and I own the process, and I can say, hey, bummer that I spent, as much as we did on R and D developing that tool, but now I actually get to decommission my tool and put in a better tool, and like that's a actually a good outcome. that's positive Ingeno: Yeah. Jeff Cripe: for the business. It's it's a bummer that you wasted the money on the RD, but hey, you got value out of it while you had it, now you have a better tool. And my company's fine. But if I'm just selling the tool, I'm dead. So we also Ingeno: Right. Jeff Cripe: just thought about how do you future proof yourself from these models that are so rapidly improving in so many different areas? And we thought it was really important to own the client and own the output. Ingeno: Makes a hundred percent. it makes sense. but I think you said you started in twenty twenty one, right? So at that time, cloud and chat GPT. Okay. Jeff Cripe: I met my co I met my co founder in twenty twenty one. We started working on Procode in twenty twenty four. Ingeno: Okay, so LLMs were around at that time because otherwise I would have said you're you're a visionary thinking that Cloud might replace you down the line, but okay, I understand. all right. so we we talked about a decision that was a real win for you. What would be a decision if you could you would take back? Jeff Cripe: I mean it's not huge. It probably costs us three months of of like dev time, but we just we built our products out of order. We should we built the provider app first. well sorry. We built the coding engine first. That was the correct thing to do. And then we built the provider app and then we built Procollect. We should have built that the other way around. What we didn't appreciate at the time was the way that we were actually gonna access that information in a scalable way. It made more sense to build the other thing first and then power pro our provider app off of all of that data that you were bringing into our system and and validating and structuring. So that probably cost us three months, which was a bummer. But you know, in startup land, like three months is a lot of time. Ingeno: Mm-hmm. Jeff Cripe: so yeah, probably that. Ingeno: But learning is also learning is also part of the journey, I guess, right? Jeff Cripe: Yeah, it is. Ingeno: to another founder or CTO building health tech software. Listening to this, given let's say the complexity of integrations, compliance, the AI pressure, the hiring, what would be your honest advice on where to start, how to be successful? Jeff Cripe: look, I think it's the most interesting time Ever to start a company, Ingeno: Mm-hmm. Jeff Cripe: in so in so many ways. And and my only regret, I started my first company pretty young. I started my first company at 25. Like I wish I'd started a company in I in you know, high school or college. It probably would have failed. But like the the refs that you get from just starting are super Ingeno: Mm-hmm. Jeff Cripe: valuable. and I think early the reason I didn't start a company earlier is I was limited, I think, by financial acumen. and two I I was limited by my own technical capabilities or lack thereof. And you sort of have to organize your ability to build something around both the things that you're good at and the things that you can't do at all. and those were two big barriers for me. You know, financial ability, to to build and forecast a business and understand a PL at the time when I was twenty years old and the ability to just write code. I didn't really have either of those things. Ingeno: Mm-hmm. Jeff Cripe: it is so easy with AI to solve for those things now. You can build an MVP and get it in a market without an engineer. y you can build a financial model in thirty seconds if you can give it all the right inputs and and it's easy for me to think through the right inputs. It was hard for me to build a model. you know I a lot of my friends went into banking and consulting. Like I didn't like I I didn't, you know, spend three years living inside of Excel. so I that my biggest you know advice and recommendation and it always is to young founders is just start as early as you can and like don't don't you know worry if it fails. you're gonna get so much value out of it failing and so many most of the really successful people you know have got some failures in their past and have probably Ingeno: Mm. Jeff Cripe: in some ways, you know, that's led to led to the success they've had. And I think the other advice I would give is just try to figure out what your superpowers are. I think it's really v valuable that the the CEO of the business spend time on the things that they uniquely can give to the business. And you have to wear a million different hats, but like I I really think you have to organize your schedule and your priorities and your time allocation around the things that you have to own and you should own because those are your superpowers. And I didn't know what those were early on. I had to discover them along the way and now I do. And Ingeno: Mm-hmm. Jeff Cripe: I'm, the second time around I'm a lot better at organizing what I spend my time on and who I ha hire around me to build a complimentary and really effective team. but I think it's worth kind of as you build your career early on, trying to figure out, you know, those things for yourself and then apply those as you as you move forward. Ingeno: I love that. All right, Jeff, before that you go, quick lightning round of questions. First thing that comes to mind, no rethinking. What would you say is the most overrated thing in healthcare technology right now? Jeff Cripe: Gosh. Hard to think. I, as much as I said I, I was into the the the sort of physician in your pocket idea. I don't know how widely penetrated that is or how that really compares to you know an actual physician. So I I I guess my mind goes there, but I say that without a lot of confidence or knowledge or experience. I Ingeno: Right. Jeff Cripe: ha I I let me put it this way, I ha mo none of my friends have ever told me like, hey, I consulted an AI instead of going to a physician. It might be I consulted an AI and And then went to my physician, but I've never heard them say I didn't go in the f I didn't ultimately go. Ingeno: Yeah, yeah. But may I guess many people are asking chat GPT these days. I don't know. Yeah, yeah, but Jeff Cripe: For sh for sure. But I still feel like they end up in a doctor's office. Ingeno: they they should, yeah, I agree, I agree. Jeff Cripe: And unless they're just figuring out their own their o unless they're just figuring out their own peptide stacks. I do know some people doing that. Ingeno: the one tool or system your team could not run without. Jeff Cripe: gosh. Claude, honestly, Claude. Like Ingeno: Yeah. Jeff Cripe: it's w yeah, if if you just look at the scale of our output with this relative to the size of our engineering team, it s wouldn't be possible two years ago. Ingeno: Yeah. Crazy, right? It it goes so fast. I guess what your answer is going to be. I I think I know what your answer is going to be, but what would be the best engineering or product team you've ever seen and what made them stand up? Jeff Cripe: That I've seen or been a part of. Ingeno: Either. Jeff Cripe: I would actually say perplexity. so somebody somebody I'm close with is is very senior in that business and just the velocity of the products that they're shipping, their ability to pivot from exclusively a consumer search app into like an enterprise perplexity computer tool, the the the browser, the velocity of partnerships that they are shipping to drive distribution, it's insane. it genuinely Ingeno: Mm. Jeff Cripe: is insane and I and I know that so much of what they do is based on asynchronous decision making and that's really culturally important to them. And I think it's tied to that philosophy. But I would actually say honestly perplexity. It's it's like pretty unbelievable what they're doing. Ingeno: Nice. what do great health tech leaders do that the average ones don't? Jeff Cripe: Gosh. I I think it's probably at least in our neck of the woods, I I think it's just like persistence and the ability to battle constant edge cases and not get discouraged by it. It's really a hard place to build and automate. there's just differences between every EHR, how different clients use every single EHR. it and those small problems just get reflected into much larger problems when you're trying to deploy software at scale. and I I think it's probably just like persistence and doggedness and a refusal to give up and settle for you know, a mediocre product or eighty percent of a solution. so I I think probably at least, you know, I probably that. Ingeno: Right. Anyway, everybody else has the same challenge, right? It's it's it's it's kinda inherent to healthcare, I guess. Yeah. Jeff Cripe: Mm-hmm. Ingeno: if you were not an entrepreneur in software engineering and healthcare, what would you be doing? Jeff Cripe: I lo I have a buddy who, you know, did well early on and now spends a lot of his time doing wildlife photography. and Ingeno: Mm-hmm. Jeff Cripe: it looks amazing. And I I that's always been like a dream of mine is just to like travel for a long time and like you know, be in incredible places and see incredible animals and and take their pictures. So I I might do like wildlife photography. Ingeno: Nice, nice. All right, Jeff Krepp, this was a genuinely great conversation. Thank you so much for taking the time to discuss with me. I really appreciated how candid and honest you were about what you guys are building, your industry, the challenges, the good things, the less good things. and this is exactly what this show is for. So tell me, Jeff, where can our listeners find you or follow what you guys are building? Jeff Cripe: you can check us out. So we've two different websites, triprocode.com and and octusconsulting dot com. that's the company that we acquired. S same company. We ultimately have to merge the brands together, but can find us there and I don't have much of a presence on social media. You can find me on LinkedIn. Ingeno: All right. Thanks everyone for listening to Digital Leaders Club. If you're a health tech leader dealing with any of the challenges Jeff and I talked about today, I'd love to continue this conversation with you. You can find us at ingeno.ca. Until next time. Thanks again, Jeff. Jeff Cripe: Thanks.