Episode 24: Rohan D'Souza, CEO of Avante

Speaker A: Name is Norm Bolsky. Your host, phenomenal hires that you've helped place because that's, that's the other thing, like when it comes to enterprise sales. This is the Digital Health Heavyweights Podcast. Honor to be here and it's been

Speaker B: fun to work with you over the years. We watch our dollars, we watch our. But people win with us, and that's important.

Speaker A: Hello and welcome back to the Digital Health Heavyweights Podcast with your host, Norm Bolsky. And today we are thrilled to be joined by co founder and CEO of Avante, Rohan d'. Souza. Rohan, welcome to the pod.

Speaker B: Thanks, Norm. Uh, I don't have my boxing gloves on.

Speaker A: It's, uh, okay. I know you're fighting rising health care costs along with us, so.

Speaker B: I, I sure am. I brought, I brought some gray hairs to go along with your boxing gloves.

Speaker A: I like it.

Speaker B: Um, uh, thanks for, thanks for having me. I'm really looking forward to this.

Speaker A: Yeah, no, I've been as well. So, you know, for all of our viewers and listeners that aren't as familiar with Rohan, Rohan is CEO and founder of Avante, as well as earlier in his career spent time at Eclinical Works, Kensci, Olive in healthcare data and technology. So today, Avante is helping employers use AI to better understand their benefits programs, improve employee engagement, and make more informed decisions about healthcare spending. So we're excited to have you on and tell us all about Ivante and application of AI and health and benefits.

Speaker B: You bet. Topic du jour. Right, everyone? Everyone wants to hear about AI.

Speaker A: Certainly, yes. Uh, it's all the rage. So you've spent most, if not all of your career in healthcare data analytics and AI. Looking back, what originally drew you to the space?

Speaker B: Yeah, thanks for the question. Uh, I accidentally fell into it, Norm, going all the way back. Um, I'm a first generation immigrant. I came to this country on a sports scholarship and I went to school in India and like most Indian kids, went through a pretty rigorous academic, um, system. I was never a math person, but I was a very curious science person in general. Sort of like gravitated, you know, to this space in and around data. And as I look back, you know, I've spent my entire career at the seam where healthcare data meets the people who have to act on it, whether it's the provider, whether it's the payer, now obviously the employer. And as I was thinking about my last sort of 18 years on the provider side, I was like, the last place that data just hasn't reached is the employer. And the employer now especially coming out of this post Covid world is really the thing that is funding for maybe a hot take here, basically funding the entire United States healthcare system.

Speaker A: Because as multiple times on this pod, like paying 262% of Medicare on average.

Speaker B: Yeah.

Speaker A: American employer, and thus the American employee is floating the health care system period, 100%.

Speaker B: 100%. And those consolidations on the supply side are just going to continue to create a crazy amount of pressure for the employer. I leaned on my previous background, right. So I started off, as you mentioned in the introduction, on the provider side at EClinical Works, for those people who don't know, but of the largest ambulatory medical records vendors, I was super fortunate very early on coming into this. Um, I got thrown at the biggest and largest implementations. And as I was going through that, there's a joke on the provider side that they say most EMR companies are basically revenue cycle management companies because that's just how the system operates. And so I was like really, really seeking how do we take all this clinical data that's being captured and make sense of it. They ultimately just became revenue cycle programs. And then I left to this really awesome startup which pulled me out here to Seattle that had pioneered this interpretable machine learning concept that was telling people where the problems might be and what the models were forecasting and try to apply that to population health. It eventually became, you know, R1, purchasing it as part of Providence and became a rev cycle program again. And then it's just like that same

Speaker A: door back to rev cycle.

Speaker B: It always does, man. It's like keeps on pulling you back there. And then the same thing kept nagging me. You know, it was like data existed but the humans couldn't act on it in time. And then benefits is this place where the employer spends the majority of their money, has the least amount of visibility. People are completely confused. Ah, it's caused major, major friction. And so that was the pull. It's like, let's go figure this thing out.

Speaker A: Yeah, there's certainly a lack of, you know, having the UI that a lot of other consumer shopping experiences have. And it feels like we're in the middle Ages still. So it's a really good time to apply new technology to old problems and persisting problems. So at eclinical Works and later at Kensci, you're helping providers better understand populations and risk. How did those experiences shape the way you think about healthcare today?

Speaker B: Yeah, so first of all, AI is challenging the very existence of traditional software stacks. So keep that in Mind for your listeners also, you know, as everything that we have known as incumbents are being relegated down in the stack. But as I look back on what the provider side taught me is prediction without a channel to act on is just like a nicer report. I had this mentor who always would challenge me. He's like okay Rohan, this is a really awesome thing that you're telling me but does it pass the sowat test? And the sowat ah test is like okay, you've identified a large percent of the population does not know that. You know I'll name some names here like lira is a wonderful EAP provider but the majority of people don't know it exists. Uh, you're telling me that Rohan, so what, how do we drive that change? Right, so at eclinical works it was the record only matters if it changes the next clinical action. So our whole thing with population health is it's got to show up in the right moment at the right time. Which means when the patient is sitting with the provider having how is the provider nudged that hey, Rohan is at risk for pre diabetes and you need to remind Rohan that X, Y or Z is available to them and they should take action on it. The same thing applies here, except the employer does not have the system internally to infinitely scale. And AI just provides this amazing now opportunity where you can take the experience of the person inside of the organization, their entire ecosystem and infinitely scale it in an n of 1. The beauty about it is that AI is providing for us is it is a true first time headless opportunity. It's like show up wherever work happens. If a person chooses to communicate primarily over slack, you just got to figure out a way that you show up in slack and make it happen.

Speaker A: Yeah, you don't want people changing their behaviors. You want to go where they're already their time. Yeah, exactly. And throughout your career you've helped build products for providers, health systems, health plans. Um, you know, looking back, what lessons uh, have proven most transferable as you've shifted to you know, self insured employers and benefits these last year and a half, two years.

Speaker B: Yeah. You know, so if I were to like sum it down into three big things. Number one, building a system of trust. You have to like, number one is like earn trust where the narrow wedge before you sell the platform. Right.

Speaker A: Makes sense.

Speaker B: So people think of us initially as like oh, you're selling ticket deflection software. We're not selling ticket deflection software. Right. You will get ticket deflection as a product of people Trusting the AI because they're just going to go to the AI first and as a product of it, you're just going to get a whole bunch of tickets that are deflected. Right.

Speaker A: Ah.

Speaker B: So I tell people this all the time. If we're not going to have an honest conversation about the trust that's going to change inside of your organization selfishly for us through the wedge of our product, then we're actually not a good fit for you. Right. So one, earn the trust with a narrow edge. Second, I already hinted at this. Meet people inside the tools they already use. Do not introduce another application. There's already enough. Right, Right. HRIS carrier portal, TPA portal, internal call center. Yeah. I mean, just like everything, right. Everybody has a solution today. It's like you got to figure out a way that you show up in the tools that these people already use.

Speaker A: Yep.

Speaker B: And then finally, just never ship an analysis a human can't act on. If it is something that is not actionable, it is considered a vanity metric. It should not pass the bar for you to ship this out.

Speaker A: You.

Speaker B: And you sit back. Right. You take that.

Speaker A: You know, I, Yeah. So where you're going. But I want to make sure any listener that isn't like overly technical.

Speaker B: I mean, like the, uh, example, like we're coming off of we're leaving Mental health awareness Month. And there's a lot of insights that you can glean from conversations that are occurring. Several of them are immediately actionable. For example, your EAP documentation might not highlight the fact that it is available for dependence on the plan. You might be thinking in your head that, oh, this is only available to me as an employee. That's a very tangible, tactical thing that you can drive today, right now. Right. Versus like, oh, you know, the network associated with your EAP provider turns out is not the highest quality. It's not something you can action on immediately. You can have a conversation about it.

Speaker A: Contract, um, renewal or whatever.

Speaker B: Exactly. Right. You got to pull a whole bunch of people in. And so a simple exercise I tell. I mean, there's a whiteboard here to our left and we have two circles. It's like obsess about the things that you can control on. Be aware of the things that you can influence. Don't fret about the things that you have no control over. Right, True. So like, very tactically, like, what are the things that you can solve for today? Today you can solve for updating your documentation to tell people that your EAP solution is available for dependence.

Speaker A: Yeah.

Speaker B: Simple as that. Right. And test it. Test it in 30 days.

Speaker A: Yeah, I like that. Really, really cool. Uh, how are employers different than provider groups or health plans?

Speaker B: Well, the beauty about employers, especially for the self funded ones, I mean, whether you like it or not, you are functioning as an insurance company. Right.

Speaker A: So how to benefits is running their own health plan.

Speaker B: It's exactly right.

Speaker A: And then they have to do everything else on their do list, but unfortunately

Speaker B: they're a cost center. And cost centers are under resource. Yes, exactly, exactly. So, uh, I think the difference is you get to straddle the employee versus the patient versus the member. And so you get to see this like perfect harmony of like, okay, who's the individual as an employee? And then they enter the system as a patient, either through the lens of being an employee or being a member of the plan. The health system always looked at a person as a patient. The employer looks at them as like, hey, they might enroll into our plan and then eventually they might be consumers as patients, but really to us they're an employee. And that creates this really interesting communications and sort of population health and population employee strategy that's very different than the health system side of the house.

Speaker A: Yeah. And I would say this, what I think makes a really cool, like, even next layer to that is, let's just be honest, health systems make more money when people are sick, period. Same thing with health plans, same thing with PBMs. You know, it's the incumbents that make a lot of money from chronically ill people. The employer does not want their employees chronically ill for two reasons. Number one, they got to pay for them. And all of those, you know, chronic diseases, very, very large costs in the future, if not immediately. And to like, they want you productive, they want you feeling your best, they want you performing your best. So I would argue employers are the only big entity in all of health care that is actually completely incentivized to keep their employees healthy, happy and working. And I think that is a really cool insight when starting to work in this employer market is their incentives are more aligned than, I would argue anyone else in the system to the actual patient and their long term quality of life.

Speaker B: 100%. Yeah. And that's physical, mental, financial. The example I love giving people is somebody, uh, finds out they're pregnant, it's a really exciting moment in their life, they're about to graduate from being an employee to a patient, and then also have to deal with this very terrifying thing called disability. And they're like, wait a second, I'm not disabled. I'M giving birth. But, uh, that's how the system treats them, right? The system says is like, oh, you have to go on short term disability because now according to the rule of the law, you are considered disabled. And you're also a, uh, patient and you're also an employee. Think about everything that's going on.

Speaker A: Large bill from the health system that you might not understand or know what you should pay and what you don't.

Speaker B: 100%. 100%, right. So, yeah, it's super exciting time of how we can stitch together all of these, you know, available content and contextualize it for us all as consumers.

Speaker A: I have a funny story, actually. Our second kid, we're in the hospital, Molly has just given birth, and she gets an email from her corporate HR department saying, hey, uh, since you're a part time employee, you're not given maternity leave or anything like that. Like, you have to be at work tomorrow or you don't have a job here. And my wife just was like, well, I told my bosses, like, yeah, I'm not gonna be there tomorrow. Uh, I just had a baby. They figured it out on the back end. But, like, funny how callous, uh, some outreach can be at, like, very, very sensitive time. Luckily, Molly took it in stride with the laugh. But m, you know, it is pretty shocking what, you know, comes out of, uh, corporate, you know, HR and benefits departments, if you don't have layers in between that are, you know, making sure everything is done at a high level.

Speaker B: 100.

Speaker A: So, you know, that's in a very pre AI world. So, you know, I think we're putting out now can really soften, you know, some of those things with some empathy. Yeah, really, really good stuff. So what ultimately convinced you that employee benefits, not hospitals, health systems or payers were the next problem worth solving?

Speaker B: Well, the sort of, like, obvious thing, Norm, is like, it's the largest pool of health care spend with the least intelligence sitting on top of it right now. Right. I just made the economics work for the first time. You know, this. People have always been attracted to the employer side of the house because, you know, if this concept of self funding, you know, 65% or maybe even larger than that now are self funded, they're running their own insurance programs. And we came off a little bit of this high with COVID where there was a crazy amount of purchasing that occurred while employers were trying to figure out, you know, what do employees actually need. But, you know, just like rough math, employers are funding roughly 1.5 to 1.7 trillion of the U.S. health care economy, most are still flying blind and it's not because they're incapable off. They've been around a system that has been sort of protective and made it really hard. Thank God for folks like Cuban and others now that are making such a big noise about it.

Speaker A: It was just on the pod.

Speaker B: Oh, amazing. Wow. I feel honored that um, you'll be

Speaker A: right wedged in there, you and Mark, back to back.

Speaker B: Um, but you know, folks like that who are rightfully calling out that, hey, you're sitting on this gold mine of information. Uh, but there's as most information asymmetry problems. You have a data silo problem that really does have to be stitched together. That's where sort of I saw the opportunity with our early team and we said we're solving an information asymmetry problem. We have the amazing thing of we're building a company at a time when arguably the greatest technology ever created for our generation and God knows for how long is there for us where we have no legacy. Uh, there's a ton of incumbents, you know, very large consulting, uh, companies that are, you know, generating upwards of 30 billion of human powered consulting margin doing annual work. That's kind of up for grabs at this point. Right. And then the CAA disclosures rules turn these lights on for employers beyond the lawsuits that are occurring about what these dollars that they're spending are actually getting put towards. And can you recoup some of that internally? And I'm a software guy, right. So I happen to be in healthcare, but at the end of the day I look at things in first principles of what can software do to unlock. And I was realizing there was no software that existed within employers. It was mostly cobbled together with reports that they were receiving or anecdotes going

Speaker A: into quarterly self reported data.

Speaker B: Yeah, exactly. That were your incentives were not aligned. Right. It's the vendor telling you that here's the 7x ROI when all we really have is just a sliver of data. And my eyes really started to open up when I had the opportunity to spend some time with some early adopter, very large jumbos and said hey, we could make a real dent in a flywheel that can spin inside of the employer and not come from the outside. That was enough to say like hey, we should go after this problem.

Speaker A: That's awesome. Yeah, I love the conviction. Uh, for listeners hearing about Ivante for the first time, what does the company do and what problems are you solving?

Speaker B: Yeah. So Avante is the AI native benefits we Call it the AI native intelligence layer for employer health benefits and beyond health. Uh, now we see navigation is really the door and the intelligence layers, the building a lot of the early messaging that AI got a lot of traction is the chat layer on the top. We were never building the chat layer. We were building the chassis, the harness, the knowledge base underneath it.

Speaker A: And that's where does that chat sit?

Speaker B: So the beauty about the chat is the chat sits wherever work happens. You know, super tactically, we built a really, really great API, uh, with all of the integration hooks that you want. Right. We have MCP on the front end, we have a 2A on the back end. So this is a model context protocol on the front to allow rest APIs to convert in language that agents understand. And then how do you take an agent and talk to another agent? So can Carly, our agent, actually talk to a United agent that an employee is asking a question about a prior authorization? Like we've built all of those harnesses behind the scenes and so quickly for us is like we looked at the world, like I said, at the intersection of the employee and the employer. Carly, named after hm, my amazing co founder, Dr. Carly is our employee facing agent. We allow people to call Carly whatever you want, but we were like, every organization should have somebody like Dr. Carly. She's an AI ethicist, she does AI research, she teaches at Duke. Super passionate about this product.

Speaker A: Like homage to your person.

Speaker B: Yeah. For what it's worth, she hates it, but we all love it.

Speaker A: Yeah.

Speaker B: Because now she hears her name being called out everywhere. Uh, but so Carly is the thing that's handling, you know, at many, uh, of our organizations, like at Zscaler, it's processed 14,000 employee conversations. If you know, just do the math on that. That was 14,000 unique moments that mattered in people's lives. That if you needed an HR department to surface that it would have to be 27 full time people. That's just not going to happen. Right. So baselines are getting completely changed and then Ava is just. Nothing really creative there. It's short for Avante Virtual Assistant. Ava is the agent that sits across the entire platform. So Ava surfaces things like, hey, your SPD document and your benefits guide are completely out of whack. Uh, the SPD is saying this and your benefits guide is saying this. You're seeing a huge traffic in 401k questions coming in in April and there's clearly something going on. You might want to look at your comm strategy and by the way, your GLP1 utilization is up 31% this month. Here's how you should think about that going into a renewal conversation. It's a whole new way of thinking about roi analytics insights that closes the loop or the flywheel between what people are seeing and what you as an HR leader are able to consume in this really enriched data format.

Speaker A: Love that employers spend millions and millions of dollars each. You know, obviously billions collectively. Trillions if you add all up on uh, health benefits every year. Yet many still struggle to answer basic questions about utilization outcomes, roi. Why is that visibility gap persisted so long and you know, what's the opportunity now to bridge those gaps?

Speaker B: Yeah. So you know, Bezos famously said like if it doesn't make your beer taste better, like why do it? And even though you tell employers, hey, you're running an insurance company, you're spending 100 million plus on this, it was still not the main thing. And so they came in into most of these conversations knowing that and what led to them punting or pushing it out is, well, health care data is really hard to understand. Yes, it is hard. Second, there's no consistency in the data formats. Even though they say you're supposed to get access to your data, you will get a, uh, you know, a carrier,

Speaker A: give you a claims data or like data sets.

Speaker B: Exactly. I'm talking even just a, uh, basic things around how do I know who's using what? And there is no consistency in how this information is shared. Right. So beyond the silos, there was an intimidation factor of like, hey, wait a second, I'm somebody who grew up through HR and now I have this responsibility. I never spent time inside of the healthcare setting to know what's the difference between an Ms. DRG and an APR drg, et, uh, cetera, et cetera. And so these have just become hard things that people have had to grapple with. So that's the reason why this gap has persisted for so long. Claims it with the case carriers, TPAs have your eligibility data. HRIs is telling you who's an employee, what status, blah blah, blah. And then assembling it was like a once in a year thing. Right. And you're paying a consulting company a six figure statement of work just to do like an updated report for you. Until AI. Producing that intelligence continuously was just, it was way too expensive, it could not be done. The math didn't make sense. Uh, with AI now that has totally changed.

Speaker A: Yeah, I like that. What are you seeing in the data around like where employers are probably spending more than they have to and like where are some of these efficiencies going to come from. So hopefully in a handful of years like 9, 10% renewals will not be the, uh, expected increase because we're actually able to, you know, make things more efficient in terms of dollars being spent to what value you're getting back.

Speaker B: Yeah, I mean, look, we're in the early innings of fully unlocking this data set and the insights that you can glean from it. A little bit of a grim sort of outlook, I think. What's happening on the supply side, Norm, with so much consolidation and employers picking up the bill, I don't see a path to the 8 to 9% changing anytime soon. And I look at things like an 18 month cycle because that's just the force of what AI is forcing you to do. Um, you know, the supply side is being constrained too much. What employers can do is look at where they have overlap and similarities in their design. And for the first time now, if you have all your contracts data, you have all your utilization data, you have all your design data coming in and you have the types of questions people are asking, you really have an honest fingerprint beyond the footprint of the data that you have. Um, on why do I have 13 different telemedicine providers? I don't need it, right. Can I use some sort of a router to take me to the right place in the disease states that are needed and get rid of a lot of the bloat that I have in my system? Previously you never had the ability to interrogate the depths of your contract to say like, oh, wait A second, my ASO contract actually has like 24, 7 virtual primary care built in into it and I just didn't know it existed. So I'm punting people in a different direction that now I don't need to do. So I think it'll be a real wake up call for what? You know, purchasers have been pushed from the outside when they themselves internally have access to the data and can go back to renewal conversations now and say, hey, I don't need it. Right. Admin fees are going to get cut significantly.

Speaker A: I still go back to um, like the broker side. Which side?

Speaker B: Yeah, broker, broker, health plan, primarily on the brokerage and on the carrier side is where you're going to likely see a significant amount of changes. You now have the ability to have a very honest conversation armed with your data in there. Right. By the way, none of this can happen without a system on trust. Right. You have to have the system of trust that is built where people are actually using Your products. You're able to interrogate the quality of all of these point solutions that are getting put in place based on your population, not some um, hypothetical from the outside.

Speaker A: I uh, like it. And you mentioned like there's certain things that brokers and consultants are doing today or performing as you know, a fee based work that AI can handle. Can you kind of walk us through like which things can be outsourced with AI uh and drive some significant value back into the employer in a bank accounts?

Speaker B: Yeah, I mean like I've heard crazy things around, hey, the you know, multiple time a year pulse survey that needs to get issued is completely gone now. Right. That's a significant, you know, one time engagement that has to be done in sort of like an old school survey style model that now with people talking to an AI you don't need anymore. Right?

Speaker A: Yeah.

Speaker B: And look, I think brokers and consultants themselves are going through this incredible modernization of their own very existence. And I've seen this uh, with the ones that work very closely with us, we've publicly shared the ones that have worked with us, they have accepted the fact that they need to disrupt themselves or be disrupted. Right. So many of them are going in and say like, hey, uh, as a way that we engage with you, the customer, we're going to be very AI leaning and AI forward, which means you're deploying Avante out. Our consulting team have access to it, we're working in it together with you so we can do more real time analysis. That's a pretty big change to an old school way of like hey, we're going to go in the background, do a bunch of analysis and come to you. Right. And so the forward leaning brokers and consultants have had to do that and it's probably going to unlock new forms of revenue for them. Right. So which is I think what's making them very excited about the potential of AI as they bring this out to their customers.

Speaker A: Yeah. I think whether it's benefits internally or consultants externally, if you're able to do a lot using AI, eliminate the need for manual tedious tasks, it should elevate people to work top of license.

Speaker B: Yeah, nobody wants to sit here, Norm and gets excited from the 80% of people who come ask you a question of like, hey, where's my insurance card and can you remind me if this thing's available? Like they don't. Right. A consultant really wants to be a consultant. They truly want to get a sort of like feeling of what's available in market and then use real customer Data to inform whether what they're seeing in market makes sense for this customer. In many cases it might not. Right now you actually have the ability to do true consulting.

Speaker A: Yep. This episode was brought to you by the digital Health team, Direct Recruiters as well as the MVP Growth Partners team. We are really excited on the direct recruiter side to work with companies including Livongo, Teladoc, uh, Hinch Health, Lira, Virta, Progeny, Carom, Rightway and many other great companies from MVP Growth Partners. We're really excited to invest in early stage digital health companies. Our first is one imaging which we couldn't be more excited about and posterity being second and many more to come. Thanks for watching and listening. And back to our episode. Healthcare benefits are one of the largest investments employer make outside of payroll. Usually number two or three. Like it's been said that Ford spends more money on benefits than steel for their cars. Why do you think organizations have historically had better visibility in almost every other major expense category other than healthcare? Besides the things we already discussed?

Speaker B: So a couple of reasons. One, most other industries have been very. Let's pull the steel analogy for a second, right. It's a pretty well defined ERP esque supply chain. Everything has a unit number, it has a product sku, it has an id, it has a cost allocated to it. Healthcare unfortunately has never operated that way. The ontologies and the intentionally, I don't know if it's intentional. I think we've gotten creative. I'm giving the benefit of the doubt that human beings are good human beings here. And so I think because the complexity of coding and upcoding, like I come from the provider side, I spent a lot of time uh, amongst us friends here with doctors who constantly wanted to know how to code the highest level of the diagnosis code in order to get the highest reimbursement. Like let's not fool ourselves like that is happening on the provider to do

Speaker A: what's in their best interest. I think the American economy has taught us that. And ah, you know, ultimately like you show someone a comp plan like I'll show you what humans are going to do with it.

Speaker B: It's exactly right.

Speaker A: Pretty standard.

Speaker B: Yeah, exactly. And I think you know, continuing on that semantics on ontology have been hard. Uh, with healthcare, complexities in contract have been very, very hard. I think for employers what's a huge difference is it's this combination of internal and external. There's data that you have that is internal to your organization and you are dependent on a ton on the external Side and historically m. Having those conversations with the, with the external providers has been very, very hard. And I don't mean healthcare providers, I mean just anyone in general. Right. If you wanted to get the reasoning behind, uh, a high CSAT score that, ah, you know, a diabetic care navigator is presenting to you, you've never had the ability to have an educated conversation around why, and now you do. Right now you can ask the like, tell me your inputs into my output, but you're dependent on a third party external source. Unless you're one of the few employers where you actually operated as a tpa. You know, like Cerner from back in the day ran their own internal tpa. Many don't. Right. They're dependent on the outside. So I think that's been the challenge.

Speaker A: Many benefits leaders tell us they feel overwhelmed by the number of vendors and point solutions they're managing. What are the biggest challenges you're hearing from the employers you've been talking to?

Speaker B: Yeah, look, vendor sprawl is a real, real big challenge. Uh, I start every one of my conversations. I was just at Northeast Business Group on Health and I looked the room and I told them, I was like,

Speaker A: New York, it was the day of the parade. I heard.

Speaker B: I know, it was wild, man. I'm shocked that so many people actually made it. It took me an hour and a half to even make, uh, from Grand Central. But what I tell rooms like that is y' all have a really, really hard job. And that is not lost on everybody at Avante building this.

Speaker A: Yep.

Speaker B: Our job is to like, remove the amount of like opaqueness that exists in this industry. Vendor sprawl is a really, really big problem. Right. 10 to 15 plus point solutions, each running its own sort of like dashboard and own ROI story. It's really, really hard to keep up with that. Uh, and then you've got this one to one and a half percent engagement rates and when you really look at it, you say, okay, what is it really doing for us other than saying we have it so, you know, good programs. And then what's happening is like all the really good programs are going completely unused because there's a ton of effort being put towards the 1 and 2% increase. Right.

Speaker A: And talk to us about those like, ones you put in the really good category. Like, what are those and why do you think? Even though I mean, like, look, I

Speaker B: think, I think it's proven that in the disease state worlds, like the vertical disease specific, uh, solutions that are out there when catching the employee early in their cycle and journey. It's proven to deliver a tremendous amount of outcomes and roi. Right. Take the MSK provider. Like you don't want to find the person after their back is so badly wrecked that the really outcome for them unfortunately is have to go through a pretty advanced surgery. You want to catch the people early on and in this vendor sprawl and communication sprawl, most of the people, they don't know how to ask the question. Right. So you've never been able to provide a front door for a person to come in and say, hey, my back hurts, what can I do? The only front door you've provided is my back hurts so bad that I need to find an orthopedic surgeon and in downtown Seattle. So you've completely blown past the opportunity for sword, right, Move, hinge or whoever else to come in and catch that person early on. Where the employer is paying a PEPM fee and the virtual physical therapy is available to them at no cost. Right. Those are some very tactical examples we hear over and over again. Because with all due respect, nobody's reading a benefits guide, nobody's going to the SharePoint site to look at health and well being and to pull up the document. People want to come to an AI experience today. Like they're getting in ChatGPT and Claude and Gemini, whatever it might be. Um, and that's the real power for employers to bring to their people.

Speaker A: Love that. What are some practical examples of how employers are using Avante today?

Speaker B: So I used one of the customers

Speaker A: that like, they're like, wow, this is like awesome. And you know, it's been a huge value add.

Speaker B: Yeah, I mean like one awesome example I use the pregnancy one is, you know, my spouse is pregnant and I'm weighing knee surgery and I also want to save for a house.

Speaker A: Like that's a pretty about yourself right now. You're just setting an example. Just I'm like, that's a lot going on, buddy.

Speaker B: Dude, I am, uh, no more kids in the d' Souza household. We are, we are tapped out. Um, I got a 10 and 6 year old and very happy and they keep me on my toes enough. That's great. I will vicariously live through other, uh, young couples in the neighborhood and watch their cute kids run around. But you know, if you think about that financial decision that needs to happen. Right. There's a lot going on in a person's world. They're considering some down payment for something future and they've got this situation at home and maybe they're really stressed out about this whole situation and they're weighing this like surgery that they've pushed off for a while. The ability to answer Those n of 1 personalized questions in a, uh, level and a sentiment that's unique to the individual versus push them through. Like this God awful wizard based approach of clicking buttons and then ultimately it spits out, you should pick the high deductible health plan. No, in my situation, probably not. The surest plan is probably the best thing for me to do. Even though the premiums are a little bit higher. Right. That's totally fine in the outcome. So that's a very like tactical example people are doing. And then on the employer side I have this really great example. One of our customers, she was telling me, you know, Rohan, I was with my chro in a totally non benefits related meeting and my chr asked a question and said, uh, did we see an increase in mental health utilization during the election season? And she felt so much joy that she could answer that question in 30 seconds. And I said, what happened in a world when you didn't have the platform? She said, well, first I would freak out, I'd be like, oh my God, I have no idea. Like how do I find that answer? I would put a high priority email out to all my vendors, starting with my consultant and then my EAP provider and they would all scramble because I'm an important customer to them and their whole world goes crazy. And by the time I get the response, two weeks later, I give that to my chr. My CHR is like, oh, I totally forgot I even asked you that question. It was important in that moment. Right?

Speaker A: Yeah.

Speaker B: Like that's a super tactical thing that gets unlocked right away. Right. So it's pretty remarkable to see these use cases getting unlocked every single day for us.

Speaker A: Very, very cool. What is Avante doing from a clinical perspective?

Speaker B: Yeah. So again, we are taking an extremely cautionary approach to this. Our approach has been, even though we have, you know, Dr. Carley, who's an MD, PhD.

Speaker A: Yeah.

Speaker B: We have found that our role in this very complex and crowded ecosystem for the employer is most of the employers today already have some form of a coefficient, some type of a clinical center of excellence. Right. Whether it's their carrier, whether it's a navigation provider, whoever else it might be, our job is to say like, Carly has a really, really good intent engine in that it understands. Take the back example. Carly never went in and said, rohan, you should do cat cows. Now this is a real Rohan example. My back's always messed up because I

Speaker A: Like to cycle planes.

Speaker B: Yeah, all sorts of crazy stuff, man. My six year old still wants to sit on my shoulders. Like I am just constantly under back pressure. Right. And so Carly didn't like magically go in and say, rohan, do cat cows and do like a scorpion stretch and ice your back, blah, blah, blah. Carly's like, I'm really sorry to hear that. Here's a really great thing. I see that you're covered under this high deductible health plan. I see you're employed here. I see you've not taken full advantage of your 8 free physical therapy visits. I see that you've got one available and I can set this appointment up instantaneously. Didn't suddenly go into full clinical mode. Right. And that's our role. Like we don't want to come off as yet another source now that's trying to diagnose.

Speaker A: Right. Very, very cool. As companies become increasingly global, employees often have very different benefits experiences depending on where they're located or what type of employee they are. How should employers think about creating a more consistent experience across workforce while still accounting for, you know, regional differences?

Speaker B: Yeah. So the good thing for. By the way, great question again. Another one of those things that's really, really hard to do. All of our customers are, or the majority of our customers I'd say are all global enterprises. They all picked us because we're a global solution. We're not a localized solution. We run inside of their organization. We have all of the right certifications in place and our aspiration is to be the place that is sort of like. We're like the octopus inside of the org with our tentacles in every one of these regions. Once you put the software in place, you first get a true representation of your benefits. Parity and equity across the system. Right. Healthcare is local. Right. Every community. I'm a kid who grew up in India, our healthcare system is very different than it is here. Most of the system is based on word of mouth and where your neighborhood doctor is. But at the end of the day you're still providing a primary care outlet for, for people to access. And that's a problem that exists everywhere. Right?

Speaker A: Yeah.

Speaker B: So if you were to think about, hey, I have a telemedicine virtual, uh, primary care opportunity available for US based employees. Every global enterprise wants to make that available to every single employee anywhere they are. Right. If it's in the uk, does the NHS offer some sort of virtual, uh, primary care? And in India is the private system providing that all of that content and context has to come in. And then the hard part for the AI is to make sure that it understands where the input is coming from and knows how to respond in that vernacular. So, for example, you don't use the word pension the same way in the US as you do in the uk, Right? In the us, you think of pension as true. Like, hey, I was employed here and I got a pension handed to me versus in the UK, it's treated very similar to the 401k. And so the hard part is for the AI to show. And this is why the general purpose AI does a really bad job at this, which is why people end up picking us, is we contextualize it, we localize it, we know how to provide that response.

Speaker A: Really, really cool. What should AI be used for in, um, health and benefits? And at its current stage, what shouldn't it be used for, you think?

Speaker B: Uh, well, first of all, I tell people there's a lot of FOMO with AI. I even joke. I was like, people in HR do not want to be the ones holding on to the BlackBerry while the world has moved to the iPhone. Because HR and Benefit specifically has always looked at as this place that's been the laggard. But I think what's holding HR leaders back right now is FOMU over fomo. And the FOMO is the fear of messing up. So the fear of missing out is

Speaker A: further for, you know.

Speaker B: Exactly. But, you know, I gotta be careful, Norm. I mean, my kids might be listening to this episode.

Speaker A: That's.

Speaker B: And what I tell people to get over the FOMO is look at a workflow that you have today and ask yourself, can I supercharge this with AI? So I tell people you were Tony Stark and if one day I handed you the Iron man suit, what would you do with it? And the first thing that most of these people would say is like, I'm still in the Iron man suit. Right. I didn't let the suit just sort of fly on its own. And so augmenting workflows, things that come to mind, you get 100 emails a day from employees asking you questions on your benefits. Can you respond to 98% of those with an AI assist? Which means, like, can you put the suit on? And can the suit let you respond to these in real time? Right. That's a great example of how you should be using AI today. Obviously, selfishly for us, you should put Avante in because we'll contextualize it, blah, blah, blah. But you could also just use Copilot and copy the email and paste it in. And then you might get into a little bit of trouble with your IT team and say, hey, wait a second, you should not have done that. Which is again another case for why you would pick Avante. Those are the places that you should be using AI today. Where you should not be comes down to what is the risk tolerance and stance that you're willing to take? Right. Can AI actually do a really good job of triaging a medical image or your medical records and point to a right benefit available to you? I would say as an AI builder, absolutely yes. Internally though, you have to be willing to accept your data privacy agreements, your willingness to take on risk, your insurance policies, blah, blah, blah. And so I would recommend every HR leader skirt to the absolute extreme without crossing the line. Right. So that's what I would suggest with

Speaker A: where technology is moving. How do you see benefits leaders and total rewards leaders day to day changing, like what are they going to be spending more time on? What are they going to be spending less time on?

Speaker B: Yeah, first of all, their jobs are completely changed. Our earliest adopters have completely rewritten their job description.

Speaker A: Mhm.

Speaker B: I will also say that our earliest adopters all got promotions. The majority of them became uh, heads of AI enablement inside of their organizations. They got operations handed to them. And so what is to say, right, the best of the best actually end up working more? Norm, they don't end up working less, but they work differently. Because what ended up happening is AI took away the mundane, it completely took away the things that they got randomly slacked about, randomly emailed about, that just prevented them from having time to strategically, strategically think, hey, what would happen if I introduced a $0 copay plan into my organization versus now you're not scrambling one week before a renewal. And so they're actually working a lot more, but they're working in tandem with almost like a digital twin of theirs, in our case Ava, that infinitely expands and extends for them. And with this agentic world that we've moved to, Ava's now moving it into a sort of like self nudging, learning thing. Like Ava's telling our earliest employer customers like, hey, you're up for a renewal in 35 days and your adoption rate on this particular point solution has been abysmal. And so you're not going to lose anything by actually not renewing. And I've given the full thorough assessment. Right. In healthcare we used to have the. I don't know if you're familiar with the S bar, but we write S bars, right Situation Background assessment recommendation. Here's a situation, 35 day renewal. Here's the background, one and a half percent. Here's the assessment. I ran a full a B test. Here's the recommendation. Right. Your consultant and broker is not going to have the time or the ability to do that. Also their incentives might be a little bit different. Right. They make money off of bringing something over to you. And so that creates this whole new way of the global benefits leader and their teams job to completely change.

Speaker A: Yeah. I'll ask the same question for brokers and consultants.

Speaker B: Yeah. Similarly man, I think you have to go into these conversations knowing your customers have access to this information and so to be prepared for it rather than assume that they don't have it. The challenge that uh, the brokers and the consultants are going to have is um, the level of understanding of AI on both sides of the aisle. You're going to have some that are very forward leaning on the broker and consulting side that are spending a lot of time in Claude or their internal AI tools or maybe using it as part of our deployments that are working with customers. It's sort of like a two by two really high AI knowledge on the consultant side, really high AI knowledge on the employer side. And it's create this nice bring out your boxing glove match and then you have the reverse of it. Right. What I'm seeing on the brokers and consultants we work with very closely is none of them are running away from it, they're all running towards it and they are really wanting to learn how to be better stewards of the platform, showing that they're innovative and truly representing the best interest of the employer. Because there's not that many places to hide. Now you just like you can't hide from, you know, something that was on page 385 of an SPD document because your employer can do that today.

Speaker A: Yep. Um, what do you think the best benefits organizations will be doing differently five years from now than they're doing today?

Speaker B: So first they're going to run like most other functions inside of their organization. To your point, the supply chain person at Ford that knew exactly when there was an issue. Forget even like a boat being trapped in the Strait of Hormuz, just like the supply chain of steel, they are finally going to get up to the same place of running every other instrumental function inside of the organization. And that comes with continuous and constantly verifying these outcomes. They'll own their own intelligence. It is truly their context instead of renting it from the outside. Right. It's Going to truly develop, uh, a core hypotheses that you will constantly test internally. And it will become this like living, breathing heart inside of the organization, just like every one of their other functions run.

Speaker A: Very cool. Love that. What's one thing the digital health industry isn't talking about today that you think it should be?

Speaker B: I think we're running away. We're not spending as much time on the tactical data sets available to truly drive, uh, information forward. What I mean by that is, you know, we say everyone can have access to claims information, but what is the quality and the level of that information that's available? Again, sounded like you spoke with Cuban and some others. I mean there's a lot that's hidden in the minutia of when you say that, hey, I can just give claims. Claims information. I think employers really need to band together as a community and demand a lot more than this. Just like over from health plans or from, I'm saying from where you're purchasing. Right. You are renting a network and you are renting a solution. You should demand a lot more clarity and depth in that data than you are receiving today. I think that's really, you know, my call to action with a lot of the employers is to really, really pull that forward.

Speaker A: Yeah. I assume your solution is only as good as the data that comes in. How, let's just say friendly or helpful. Have the health plans been, you know, who service your employer customers?

Speaker B: Yeah, they're pretty, they're pretty helpful. I think. Also, Norm, we're getting away with the fact that we're still an early up and coming platform. Um, we have earned the right to exist, but we're definitely not the, you know, one that's been around for so long that now the health plans are all saying like, what are you guys doing? Like, are you trying to build like this de identified massive data set that you're going to come after us? Uh, I was like, no, I have no interest in that. Right. I am enjoying going to battle with them when I represent certain calls with some of our largest customers that say, hey, we're working with this carrier and here's the data that they gave us. And I look at the spec and I was like, you realize there's only one diagnosis code on the spec they're giving you, right? And they're like, well, what do you mean? I was like, a CMS 1500 form has at least 24 diagnosis codes. Your average employee that has a single chronic condition actually has multiple. And the providers are billing for multiple because they want to get the highest reimbursement. So I go on these calls and I was like, hey, guys, you're cheating this employer right now. You're only giving the primary diagnosis code. I need all of it. Right. And by the way, I need the allowed rates. I need all of this stuff. And they look at me as like, oh, you must come from the health plan side or the provider side. And I was like, yes, I do. You know, let's not mess around with these guys because it's their data. Yeah, I think there needs to be more of that being talked about than just these one off hand to hand combat calls that I'm having.

Speaker A: Right, yeah. Well, it's nice you're able to like do that on behalf of your clients. You know, it takes another thing off their to do list, which is amazing. Really, really awesome stuff. Doing a, uh, startup is no easy feat, you know, comes with stress, a lot of anxiety, I'm sure. Sleepless nights, a lot of travel. How do you reduce stress and stay healthy, you know, while building a company that's been growing very, very quickly?

Speaker B: Yeah, I'm a little bit crazy. I put myself through more pain. I, uh, was an athlete. I do a lot of endurance training. I do Ironmans. Go into pain caves and put my mind at ease. But, you know, cave.

Speaker A: That sounds very uncomfortable.

Speaker B: Yeah, it is, man. It's like very grim, you know, grim place in your garage up in the northwest when the weather is super dark and uh, now it's beautiful. Like I ride my bike outside, I run outside, things like that. But, um, I like to just, you know, going into like a perpetual zone 2 kind of training mode and let my mind wander. My family's huge, you know, so I tell people I prioritize the crystal ball moments in life and give up on the rubber ball. Right. So will I make every one of my daughter's, uh, dry run routines for her gymnastics floor? No, that's a rubber ball that's going to bounce back. But when she qualifies for states, like, that's a crystal ball moment. I am not going to give up on that. Right. My team knows it, my customers know it. Like I'm prioritizing that. I find a lot of happiness and stress relief from that.

Speaker A: From that point, I like it. And then what we put into our bodies, just like the data we ingest, you know, the food we ingest certainly is a huge determining factor into the quality. Any favorite recipes or healthy meals you and your family make at home that might inspire someone to try something new.

Speaker B: Um, I'm a big Middle Eastern slash, like, Mediterranean grill cook. So anything with, like, the, you know, the base spice of a really good olive oil, some good sea salt, and then because I'm Indian, I mix it up with some, like, real legit spice, uh, that I grind on my own and always over coal and not over gas is how I like to go.

Speaker A: It's longer, but it tastes better.

Speaker B: Exactly. Man, when that, you know, the drip of the oil hits the nice, hot, uh, piece of coal, the smoke unleash into the, you know, I'm a big meat eater, so love, um, it.

Speaker A: Very cool. Well, Rohan, thank you for everything you're doing within healthcare and joining us on the Digital Health Heavyweights. And, uh, can't wait to have you back again and keep doing what you're doing and, uh, appreciate the time.

Speaker B: You bet, Norm. Thanks for having me.

Speaker A: Thank you. Thank you for joining. Please, like, comment and subscribe below if you enjoyed it.