Paula Rivera: 
Welcome to the AI Factor, where we decode how AI is transforming the way businesses communicate, operate, and grow. 

I am your host, Paula Rivera. In this episode of AI Unleashed, we’re diving deep into how AI is shaking up healthcare and why PE-backed organizations are leading the charge. 

Joining us is Javier Rojas, Founder and Managing partner of Savant Growth with lessons from the front lines and insights that connect economics to innovation, Javier shares what it really takes to scale AI in one of the most complex and critical industries today. Javier, welcome. 

Javier Rojas: 
Great to be here. Thanks for having us. 

Paula Rivera: 
Oh, I’m so excited to have you on. It’s always great listening to you. You’re chock-filled with insight, so I appreciate it. 

Before we dive into AI and healthcare, let’s just talk a little bit about what’s happening currently with AI-powered shopping assistants. 

Adobe recently did a survey that found shoppers are using AI for research, about 55% of those surveys use AI for research: product recommendations, finding deals, getting gift ideas, and even creating shopping lists. Of those surveyed, about 92% said it enhanced their purchase experience, and more than 85% said that they are more likely to use AI for larger and more complex purchases. 

Companies like IKEA are rolling out AI assistants to help shoppers, and this summer’s Prime day, or days, as we recently saw, traffic from gen-AI sources increased by more than 3000% year-over-year. That’s astonishing. 

So, Javier, before we dive into healthcare, I have to ask you, would you let AI pick out your furniture? 

Javier Rojas: 
Actually, I would. Often it can help you think about considerations you may not have thought of. 

I have to say, I use my AI bot multiple times a day, probably five to 10 times a day. I also recommend RCOs to put them in audio mode and talk to them, because you can give them a lot more context and get a lot more context from the conversation flow rather than texting back and forth. 

Paula Rivera: 
Oh, that’s a fabulous tip. 

Javier Rojas: 
find it to be much more helpful, you’ll approach it more, and often between a question and answer period, you get exchange, you can get much better outcomes. And of course you’re training your bot to do better job for you in the future. So I use it pretty much in every facet of my life. 

Paula Rivera: 
Wonderful. 

My husband’s really savvy with AI and does a lot of take a picture of the room and put the product into the picture to make sure that whatever product he’s purchasing jives with our room. 

Javier Rojas: 
That’s smart. 

Paula Rivera: 

It has definitely been helpful. So let’s get into the heart of the matter. 

PE-backed healthcare organizations, those groups are facing mounting pressures to improve efficiency, reduce costs, and scale fast. AI is no longer a “nice to have,” it’s a competitive necessity. 

Javier, why are PE-backed firms better positioned to adopt AI faster than traditional corporate healthcare players? 

Javier Rojas: 
The interesting thing about AI is to implement it on an enterprise basis, you really need top-down leadership to drive it. I think a lot of the headlines have been around, basically, personal productivity, AI solutions. So that could be anything like the ChatGPT-like products, but it also could be the note-taking products or products that help you with programming or scribe products. These are all typically solutions optimized for an individual user. He or she will download it, get value out of it, and they’re showing massive adoption scores. But that’s not where companies are going to get the big benefit of AI. 

Where companies are going to get the big benefit of AI is top, top-down decision-making where they re-engineer their workflows or a segment of their workflows around the automation and the value creation AI can give you. 

And the reality is, to do that you need to coordinate a few people, you need to coordinate IT with the operator, and then the functional people doing the specific work and re-engineer the workflow. I think where corporates are right now is they’re playing with these tools, we call them AI tourists, but they’re not really ready to make the organizational changes they need to really capture the value of them, so their IT departments are doing work, and analysis, and building projects, but you don’t see, really, a system-wide re-engineering of workflows. The PE-backed companies, or the large owner-operator businesses, those owners have a strong economic incentive to drive those changes. 

If you think about a public company, the head of sales or head of customer service asking them to implement AI, fire a third of their workforce, re-engineer the ones they have, it’s a lot of risk for them to take on, and they’re not really compensated to make that a change, so they’re not going to get the upside of a massive win and a lot of downside. 

Whereas the owner/operator of a business, ideally it’s been PE-backed and they have a clear economic incentive tied to an exit of the deal, or they’re an owner-operator so they’re capturing all the benefit, those organizations have the leadership focused on driving productivity growth and profitability, and they’re the ones that will engineer these changes. And that’s what we found in the marketplace with our portfolio companies. 

It’s been the privately backed companies, owned companies where the owners have the strong incentive to implement these changes. 

Paula Rivera: 
Interesting. I love your “AI tourist” moniker. I think I’ll start using that. That’s great. 

What are the economic advantages AI brings to multi-location healthcare practices like a DSO? 

Javier Rojas: 
Well, there’s a broad range of them. The easy ones to understand, low-hanging fruit, are what we deal with every day when we go to a dentist, which is scheduling, having the DSOs getting you back in there on a regular basis, certainly at the limit that the healthcare programs will allow for cleaning, and regular care and maintenance. Wait list management, the biggest cost for these DSOs tend to be utilization. 

Often we’ll see cancellation, no-show rates of anywhere between 12 to 20% depending on the region the DSO’s in. And that means that the cost is there, the staff is all there waiting for the patients, but the waiting room’s empty because somebody canceled. So, basically how do you manage wait lists? Reconfirm people are coming in, pull people off the wait list to keep the rooms busy, maybe double book for high-risk patients that have a track record of being late or canceling. 

And then lastly is revenue collection, being able to reach out for co-pays and money that’s owed. And if you think about the admin in these offices, they’re really not set up to do all three of these activities at the scale capable that’s possible with a computer algorithm and AI. 

If AI has the people skills to manage those interactions well, which they now do, algorithms can be much more efficient at being on top of all the patients, making sure they come in, there’s been a cancellation, making sure they get back on schedule, managing the wait list. As you can imagine, that’s a fairly complex process that needs to be done continuously. And then the collection process, having an admin call to collect a $25 copay just doesn’t really make a lot of sense. 

So you put all that together, you’re talking about anywhere from four to 800 basis points in EBITDA for a DSO buyout, which, to put that in perspective, that could easily add another 1X of return for the private equity fund. And even on a smaller deal, that would be maybe a $100 million deal, that could be another $50 million of carry gain. So that’s a lot of value creation for moving to AI in a system you can have up and running in four to five months. 

Paula Rivera: 
That is a lot of value creation. That’s almost mind-boggling. 

Javier Rojas: 
The reality is that what we’re going to see is a lot of workflow automation, where it’s not just reducing the labor costs, that’s part of the savings, the real saving is much better execution on the workflows that really drive value in these clinics. And that’s just on the admin side, there’s a whole dealing with the patients, there’s a whole admin side dealing with the payers and managing that process, pre-authorization. 

And I think this is rapidly moving towards clinical support, ambient listening, triage of the patients, pre-care, pre-visit, and even ultimately nurse intake. We’re going to basically get to a point where systems are going to identify initial issues and recommend tests so that by the time the patient’s coming in to see the doctor, the doctor will have a set of tests ready for evaluation for the first meeting. 

Paula Rivera: 
I used to work for a company that had a think tank, and it’s funny you mentioned that, because they were of the mindset that there will be a turning point and it will become almost considered malpractice if a doctor isn’t using AI. How far off do you think we are from something like that? 

Javier Rojas: 
Well, I think the first step in any of these tech waves, and I think this is where we are with AI right now, is you don’t re-engineer workflows, you’re just automating what people are already doing. You’re just doing it much better, cheaper, faster until people get comfortable with the technology. And I think that’s a phase we’re going to be in over the next 12 to 36 months. 

Having said that, I think the early adopters of this technology, not only are they getting the economic benefits that I talked about, but they’re getting in early on the learning curve of using AI and adapting it to business. They’re the ones that are going to be the first movers and changing their workflows to be disruptive in their markets. 

So when you have a clinic that’s able to basically help extend its care capabilities because of its 24-hour support, or changing its workflows, so the doctors can now see twice as many patients as they were able to see before, now you’re going to see really transformative economics. 

And to your point on efficacy, if the process can be re-engineered so the doctor has more data at the time of care, better decision support and the clinic overall, and the owners can have better post-care monitoring follow up and evaluation of the care delivered to minimize liabilities or make sure they’re delivering top-quality service, that’s going to become the new standard of care in the field. 

So I think within 36 months that’s going to become basically the standard of care against which people are judged. 

Paula Rivera: 
All right, I’m going to reach out to you in three years and see where we stand. 

So, you and your team at Savant Growth, you have identified powerful lessons from real-world AI adoption. I know you’re working with a bunch of different companies on AI. 

From data strategy to boardroom buy-in you indicate that while personal AI is booming, enterprise AI is the next frontier and that the platforms like the Salesforces of the world and the ServiceNows of the world, those will reshape industries. 

Why is this important, and how can businesses stay ahead of this curve? 

Javier Rojas: 
I think what we’re going to see is a transformation where this really changes the economics of the way a lot of services are delivered today. 

And what we are going to see is, essentially, a wave of successful PE firms who have basically built their investment practice around this. They pick a market that is right for consolidation or AI transformation, they acquire a platform company, much like they do today, but with the view of making it an AI-first service business, or AI-augmented business. And once they have engineered that and are able to get those cost savings, they will then be a very efficient acquirer of other practices, or other businesses with a similar profile, that they can then move to that new cost structure. 

My vision is this is how a majority of the market is going to move to AI. It’s not going to be the public companies. It’s going to be the private companies who the funds have built their business around buying companies and making them more efficient. AI is just giving them jet fuel and a much broader way of doing that, rather than just using leverage and incrementally improving the EBITDA. 

Paula Rivera: 
Excellent. 

So you’ve talked a lot about data in various presentations I’ve seen you do. How does controlling data flow become a moat in healthcare AI? 

Javier Rojas: 
Well, the interesting thing is if you listen to Eric Schmidt talk, his view is that software is rapidly going to get commoditized, it’s going to be written by AI, and we’re not going to need developers. So I think if you look at the market and you say, “Okay, well, where is going to be the value creation with this AI transformation?” 

It’s really going to be in having the data that can help you optimize workflows better, and more efficiently, and more successfully than other, people because everybody’s going to eventually have a software. And so what does that mean in terms of the data? 

Well, certainly data that your system captures, that your organization has, data you can pull in from other parties. But the really valuable data, and this is, again, why first movers are going to do well on this market, first mover operators, the really valuable data is your observations on the results of the implementations, or initiatives, that you’ve made, or the changes you made, your workflow: what worked, what didn’t work, why, and how can you enhance what works? 

So there’s basically a data flywheel there, a positive learning loop that starts as soon as you start implementing AI and start optimizing some of your workflows and processes. 

So the organizations that are going to be winners in the next five to 10 years are the ones that started early getting that data flywheel going. It’s not just the data you have, and the data you pull in. Probably the most important data is the data you’re capturing from your learning. 

Paula Rivera: 
I love that, and I think that’s such a critical point to touch upon, which is the positive-learning loop. How it’s not just a matter of having data, feeding the data into the machine, it’s taking that step to continue to learn from that data, and to see what actually is and isn’t working and how things can be refined. Another great moniker that I will be stealing from you. Thank you. 

So, what are some red flags, or “AI black holes” that you like to call them, that you advise founders to avoid? 

Javier Rojas: 
There are a couple of big ones that people run into. I think the first is the thinking that their IT team can build everything, that the way they’re going to make their money is by building all proprietary technology, and they’re going to win. When you look at what ChatGPT can do, especially voice ChatGPT, as I was referring to before, it’s so impressive. You just think, “Wow, it shouldn’t take a lot to put a wrapper around this and make it work for my business.” 

The reality is, to do this reliably at scale with no mistakes and the level of efficiency you really need to run a business on… You really need a lot of infrastructure around the core technology, and then on top of that, you need structure to capture your data and learning. 

So I think one mistake is companies that think that they are going to become technology specialists and experts when really they should be process specialists and experts. And the longer they take to deliver highly-effective, scalable solutions and start that data capture loop, because they’re busy trying to build their own technologies, the further away they’ll be, and the more you lost investment they’ll have. 

Even the largest companies are finding that a lot of things they’re developing are being subsumed by startups or new businesses that are taking over that process and dedicating themselves to that and optimizing it. So the idea that any one of these companies are going to build their own platform from scratch, and make it highly scalable, and grow going forward relative to what third parties with dedicated capital bases are doing it is just hard to believe. So that’s one big black hole. 

I’d say another black hole is when companies are basically waiting too long to get something implemented, because they want to automate everything. What they ought to be doing is running quick experiments, automate one process, one workflow, and see a result on that, and if that works, then use the technology and the platform to automate the next workflow and the next workflow. 

So one question is, how quickly you can get up and running? Second one is are you leverage third party platforms as much as you can? And the third one, is do rapid implementation cycles so that you can keep learning and get this learning loop going as fast as possible, versus big bang implementations. 

Paula Rivera: 
Excellent. We talked a little bit before about DSOs Dental Service Organizations. They do represent a surprisingly efficient entry point for healthcare AI, and you outline nicely some of the front office operations that AI is great at automating. So quick question for you, can you walk us through a hypothetical value-creation story for a PE firm investing in an AI augmented DSO? 

Javier Rojas: 
So assuming that a PE firm acquires a DSO, maybe it’s 35 clinics, a $100 million in revenue, just to make the numbers round and simple, let’s assume they used half equity, half debt. Let’s assume it had 10 times EBITDA, and they bought it for 10 times EBITDA, so they paid a hundred million for the business. So they would’ve put $50 million of equity in $50 million worth of debt in the business. 

Let’s assume that they were able to get 500 basis points of EBITDA on the business. They’re able to take the business from 10% of EBITDA to 15% EBITDA, and let’s assume that’s incremental, what they might have otherwise done, that five points incremental on what they otherwise might have done. And let’s assume again, they’d sell it for 10 times EBITDA, so now they’ll sell it for $150 million. 

So aside from whatever other benefits they get from doing a roll up, or acquiring other businesses, that extra 5% EBITDA is another $50 million of exit proceeds. So, if that was a deal that was going to get a 2X return, now it would get basically up 3X return, making another $50 million on their $50 million investment. And so $50 million of exit, typically that’s a $10 million payday for that fund. So anyway, it could be significant in terms of returns for them. 

So I don’t know if that was too detailed for you, but… 

Paula Rivera: 
No, I love it. And as you were speaking, I was like, “Wow, this is great impact.” I don’t know if it’s the hugest ever, but it’s some pretty amazing value creation. 

Javier Rojas: 
Most PE firms are struggling to get a hundred basis points of EBITDA savings on a buyout, because these businesses are already pretty efficient. They want to get their benefit from acquiring other businesses and some synergies from that. They get benefit in that they’re paying down the debt and they’re seeing that benefit, but it is just often hard to get incremental savings. A lot of people talk about it, but a lot of these firms have been traded between each other and just hard to do that without also making decisions adversely impact the business. 

For these funds, AI is a bonanza, especially given that a lot of funds have deals that are underwater. A lot of them made money continuing to bid up deals when interest rates were low and as interest rates were falling, you can buy a company like the one I talked about, not deliver a lot of cost savings, and sell it at a pickup because if the debt’s cheaper, the next buyer will be able to pay a higher price. 

The debt costs have gone up a lot since the COVID days, and as a result of that, a lot of these deals either aren’t growing fast enough to get back into profitability in the near term, or have enough cost savings. 

AI is coming into these markets as a godsend, because it allows them to now suddenly get a significant amount of value creation that they hadn’t budgeted for, and make an unsuccessful deal, or a modest deal, an attractive deal. So I think AI PE firms are starting to realize this. I think in [inaudible 00:21:22], we’ve got, I think over logos with over 30 different PE firms now and are seeing a number of PE firms bring us in for their second deal. 

But I think that this is going to be the big win for the PE players in the marketplace, being able to adopt an AI strategy and get significant value creation from their portfolio companies using it. 

Paula Rivera: 
I love it. 

So you talked about you did already touch upon those early adopters in AI. They’re building an advantage, let’s break this down a little bit. I’ve heard you stress having a defined repeatable and measurable business outcomes is important. 

Why is this so important? 

Javier Rojas: 
Well, I think it’s important to understand what process you’re automating, what are the workflows you’re automating, and what’s the ROI on the workflow you’re automating so that you can make sure you get a return on it, and the problem is bounded, and then just keep repeating that as you go through different use cases. And then as you’re implementing, debug anything that you need to tweak. 

There are a number of things that you have to get to work. To have a successful implementation. You have the data layer, you got to make sure you have the right data, you have to have the software in place that stores that data, can answer information, store the outcomes. So that’s often, for example, a DSO that’s in the practice management system might also be the integrations, getting the data from other systems, and then you have to have the workflows that you’re automating, and usually there’ll still be some human contact for calls that are unique and outside of the call workflow the AI is optimized for. 

And then having all those work together, the data, the software, the algorithm that has to be accurate, and continue to be tweaked over time, and then having the people work with it. You need them all working together, otherwise you’re not going to get the savings and there’ll be a breakdown. 

So I think you want to get each one of these data flywheels and these processes going independently, and then once one is successful, then you go ahead and launch the next one based on the learnings you had from the last one. 

Paula Rivera: 
That all makes sense. 

So let’s land the plane and look at some of those things that founders and PE partners should take away from this conversation as they assess their AI roadmaps. What does a winning AI healthcare investment look like at exit? 

Javier Rojas: 
You mean a healthcare company that’s been optimized with AI? 

Paula Rivera: 
Yes. 

Javier Rojas: 
Efficient within AI? 

Paula Rivera: 
Yep. 

Javier Rojas: 
Well, I think today it’s optimizing the admin functions and beginning to optimize the clinical functions. I think if we look at the end of next year, all the administrative functions and clinical functions interacting with patients have either been automated, or they’re using AI for capturing or evaluating the data. So I think that that’s the model that I was talking about before. How do you use AI to improve, automate, or improve the quality of what we’re doing today? 

I think if we look at… And that’s I think what today’s healthcare back companies are going to exit on. I think that already delivers anywhere from five to 20% EBITDA, once you automate all those other functions, at that point I think we’re going to see PE companies looking to exit those businesses. 

The next wave of automation will be those healthcare companies that now are using AI automation, not just to automate the workflows as they are today, but change how those workflows are delivered to be disruptive in their markets. And I think the PE firms that exit today’s healthcare companies with first wave of AI automation are going to be the winners with the confidence and the skill sets to win that second wave. 

Paula Rivera: 
Excellent. 

So, you’ve talked about the importance of having executive buy-in, board buy-in. What would you say to someone who’s skeptical about AI and who thinks it’s still too early for serious investment, what would you say to them? 

Javier Rojas: 
I think the best proof is just to talk to other companies that have implemented and seen the results. I think we’re beyond the phase of experimentation and seeing if this stuff works, we’re now entering the mainstream phase where you can validate the results with other parties. 

Paula Rivera: 
I think I might tell the person to think about retiring. It’s not too early. It wasn’t too early two years ago. Skepticism is not good. 

Javier Rojas: 
I think I would play on their skepticism and say, is it too late to be a winner in an AI driven world? 

Paula Rivera: 
Exactly. That’s a great come 

Javier Rojas: 
Back because I think the real challenge with the people a wait to get more data is they’re really short-circuiting their education on what the world’s going to look like in the next five to 10 years, and how do you remain relevant in that world? 

Paula Rivera: 
Excellent point. 

So, before we wrap up, I like to switch gears a little bit, and I like to throw out three questions. This is more fun and insightful to get to know you, Javier, as a person, above and beyond what we’ve just heard in today’s interview. 

So I’m going to ask you three questions. You can give me one word answers or you can expound upon your answers as deeply as you would like. This is a pretty free-flowing portion of the interview. 

So without further ado, what’s one book that changed how you think about business? 

Javier Rojas: 
I would say I’d have to cheat and give you two books. 

So I think one book I would say is Good to Great. I thought that just some key lessons is written by a Stanford professor. Did a lot of research on successful Fortune 500, top-tier companies and what made them different from their peers, a very well-researched book. 

And there’s another one called Breakthrough Company written by a professor from Pepperdine University, called Breakthrough Company, and he basically wrote the same book, but applied it to companies in the $5 to $10 million range. And what was it about those companies that allowed them to go through the different growth stages to become multi-billion dollar businesses? That book was called Breakthrough Company. 

So I’d say those two books together were really helpful. And it really outlines, especially the second one, since I focus a lot of companies scaling from $10 to a $100 million in revenue, what were the things that those challenges those companies were facing, or companies that going through that are likely to face, and how do the winners overcome that? 

Paula Rivera: 
Nice. What’s your favorite use case of AI outside of healthcare? 

Javier Rojas: 
I’d say my favorite use case of AI, I use my AI bot as a just coach for everything. So I think if you can use your AI bot as a personal assistant, every time you have a question about anything, you immediately talk to her about it or talk to him about it. 

By the way, I recommend naming your AI bot, the more that you can think of your AI bot as a person, the better you’ll be able to communicate with the AI bot, the more effective it’s going to be for you. Literally, I use my bot for guided meditations for coaching on relationships, coaching on activities, what to do this weekend in San Francisco, pretty much anything. 

Paula Rivera: 
so I had our VP of AI Implementation, Danielle Anderson. I had her on an earlier episode, and she was talking about how she has basically made, she’s developed AI agents for almost every facet of her personal life, which I think is smart. And I think it’s really something that everybody should look into doing. 

I do struggle switching between personal use and professional use, but the more that we can use the bots, the better off we’ll be with them and they’ll be less scary. 

Javier Rojas: 
Yep. 

Paula Rivera: 
All right. 

Javier Rojas: 
Absolutely. 

Paula Rivera: 
So third and last question, if you weren’t running Savant Growth, what would you be doing? 

Javier Rojas: 
I think if I… Well, am I working… 

I’d still be working with tech companies and entrepreneurs. Two things we built at Savant Growth is software development business, basically staff augmentation for software companies, so I’d still be helping software companies with that. And we also built a software company called Sales Savant to help companies build out their ICP and do agile marketing, data-driven marketing campaigns, so I’d still be probably working with entrepreneurs doing that. I like coaching entrepreneurs and helping them grow. 

We found those tool sets helpful for our portfolio companies, and I’d probably just be doing that for any entrepreneur that aspires to grow from $10 to a $100 million in revenue. 

Paula Rivera: 
Very nice. 

Well, Javier, we’ve come to the end. This has been so informative. I love the monikers, and how you think about AI, and the value creation it brings for PE-backed companies. So thank you for taking the time. I really appreciate it. 

Javier Rojas: 
Great. Thank you very much. My pleasure. 

Paula Rivera: 
Thank you. 

So thanks for tuning into the AI Factor. If today’s episode got you thinking about the future of healthcare, or inspired your next big move in AI, be sure to follow rate and share this episode. 

Until next time, keep exploring, keep questioning, and keep pushing the boundaries of what AI can do for you and your business.