Welcome And Guest Background

SPEAKER_01 0:01

Welcome to the Chalk Talk Gym Podcast, where we explore insights into healthcare that help uncover new opportunities for growth and success. I'm your host, Jim Jordan. Roughly one in ten healthcare claims get denied, and most billing teams sit on a backlog that they never truly fully work. Today's guest built a company around that problem. Rashad Patel is the founder and CEO of Arrow, an AI platform that turns denied insurance claims back into revenue. His journey started early, submitting claims for his parents' medical practice while his friends were actually at summer camp, and he continued through hedge funds and venture capital before he set out to fix the industry he grew up inside. In this conversation, we covered the cat and mouse game of denials, why AI voice bots now wait on hold so billers do not have to, and how founders think about venture capital and why the smartest AI strategy in healthcare is collaboration, not replacement. So tell me in the audience a little bit more about your journey and about yourself.

SPEAKER_00 1:16

My name is Roshin Patel, and I'm the founder and CEO of Aero, which is an AI platform for healthcare billing and revenue teams that helps them turn insurance claims into revenue. So my background's in finance and investing. I worked a few years in venture capital, working with early stage startups in healthcare. But prior to that, my whole family comes from healthcare. So my parents ran a small practice. My uncle's a surgeon, grandfather's a primary care physician. Everyone's has gone to med school except me at this point. Growing up, me and my sister were doing a lot of the billing work for my parents in their practice. And we got to see a lot of the challenges and frustration and just how much time and effort it was for them to get paid on a single insurance claim. While all my friends were at summer camp, I was submitting insurance claims and on the phone with insurance companies. And it was a nightmare experience. And I set out to try and solve this problem how do we help healthcare providers get paid fairly for the work they're doing and prevent a lot of surprise bills and unnecessary costs for patients along the way.

SPEAKER_01 2:29

So before you were at Arrow, did you also do some fintech? Did I read that you had done some fintech work?

SPEAKER_00 2:37

Yeah. At the venture fund I was at, it was a generalist fund. So they invested in everything, but I focused on fintech and healthcare. I started my career actually working in finance more on the hedge fund side. So more of a numbers person, but always was interested in the intersection of finance and healthcare, and that's where Arrow is.

Why Healthcare Payments Feel Broken

SPEAKER_01 2:57

So since you've seen the normal fintech world and you've seen healthcare, maybe share with the audience a little difference of how transactions are more complex and maybe your opinion on why it's so complex.

SPEAKER_00 3:13

Yeah, in the everyday world, when you go to Starbucks, you swipe your credit card and everything's pretty simple. You don't really think twice about it. But in healthcare, it's the complete opposite. Once you go to the doctor, the healthcare provider is responsible now for filing the insurance claim to the insurance company. So just that process of even creating and submitting the claim can take a lot of work, maybe hours of work. They have to check what insurance do you have? Is it active? Is it going to cover the service or procedure that was done? They then have to take what was done and translate it into a medical code that codifies the procedure. That's a very complex job. They send it out to the insurance company. A lot of times these are on paper or fax, so it's a very manual process. And then the insurance company can take up to 90 days to tell the healthcare provider are they going to pay or not? And very often they don't pay. Now that's when a claim is denied. And when a denial happens, that creates even more work for the healthcare provider. They have to figure out why it was denied, and that's a huge process. And then if they can't get that claim paid, a lot of times a bill gets sent to the patient. And that's why you end up getting bills in the mail you weren't expecting, like three months later. So it's a very complicated process. So does your company specialize in the physician practice part of the hospital part or both? Yeah, we focus on the physician practice. So you can think of our typical organization we work with is like regional or small, mid-size practices, could be just like a family-owned doctor's office. We don't really work with hospitals at that scale.

SPEAKER_01 5:00

So do you have a sense of how many individual insurance companies your clients are working with?

What A Denial Really Means

SPEAKER_00 5:06

Yeah, I think we work with almost all insurance companies in the country at this point. There's roughly two or three thousand, but I'd say our organizations that we work with work with about 95% of those. So your AI does what specifically? It specializes in helping the healthcare provider figure out what to do on a denied claim and get that denied claim paid. So that can involve figuring out why something was denied. So very often when the insurance company says we're going to deny this claim, they either tell you the wrong reason why they're denying it or they don't tell you at all. So you have to do a little bit of digging, usually by calling the insurance company. They don't make this data very easy to get. So part of the process is investigating and figuring out why it was denied. And then the second part is figuring out what to do next. Do you need to edit and resubmit the claim or the form? Do you need to write an appeal letter? And that process is also very time consuming.

SPEAKER_01 6:05

I owned a durable medical equipment business at one point in my career and discovered that some of the denials were at the end of quarter, which just happened to be cash flow for the insurance company. Is this an experience that you have?

SPEAKER_00 6:19

There's a lot of patterns, time of the month, or one healthcare provider told me that they started fighting a certain denial reason more, and then the payers stopped denying it for that reason. So it's almost like a game of cat and mouse where the provider is trying to, I mean, the payers incentivized to deny as many claims as possible, and the provider is responsible or incentivized to fight as much as possible. So it's a game of cat and mouse. I found that the more you fight denials, the less they happen because it creates work on the insurance company side too.

AI Voice Bots Call Payers

SPEAKER_01 6:54

And depending on how off they are, they can actually run the risk of getting into fraud, right? Which is really enabled by the system. So how does your AI tell someone like, oh, this is week three in the quarter, and this is what this insurance company tends to do, wait it out, or who do I talk to, or what's the process? What results does your software do for people?

SPEAKER_00 7:17

Like you mentioned, the healthcare providers billing team is responsible for a lot of this manual, tedious, repetitive, time-consuming work. So usually the first step to get any denial resolved is call the insurance company. And like you said, you can be on hold for an hour. And so you can spend your whole day on hold just doing three denials when you have a stack of 50 you got to get through. A huge value that we've been able to provide is we have the AI call the insurance company. So it's like an AI voice bot. It will wait on hold, have a synchronous conversation with the insurance rep and figure out why something was denied and try and collect as much information as possible. So we can call a thousand phone lines at once versus a normal medical bill that can only make one call at a time.

SPEAKER_01 8:04

So how is the insurance company person received that on the other end? Do they hang up like it's an Autobot caller or is the AI that conversationally good?

SPEAKER_00 8:15

Yeah, especially over the phone line when things are not you know crystal clear, it can be fuzzy. They can never tell that it's AI. AI has become that good. Maybe like a year or two ago, you could tell it was a robot, but the insurance rep is also trained never to hang up. They will never hang up unless they ask you, like, feel free to stay on the line to complete a brief survey. So they have a specific you know policy that they follow. But I would expect as AI voice bots become more common, payers will also adopt the same technology and they may also have AI voice bots. So we'll just have AI talking to AI at some point.

SPEAKER_01 8:51

Well, I mean, I guess the whole point of it is how do I get the reason faster and collapse that value chain, right?

SPEAKER_00 8:58

But the payer's not really incentivized to give you this information. They don't want you to fight denials, which is why they make it only accessible on the phone, hoping that the medical biller is not gonna want to call and wait on hold and fight the denial.

SPEAKER_01 9:13

So what have your customers seen in terms of results?

SPEAKER_00 9:18

Yeah, absolutely. I think there's some stats online you can look up about denial rates for a typical healthcare provider around 10, 15%. So what's great about our software is when it's implemented, that recovery rate is increased a lot. So typically the organizations we work with will see an increase in their revenue of about 7%. Not every denial is able to be overturned, but roughly 60 to 70% are. They'll also collect cash faster.

SPEAKER_01 9:46

That's what I'm just gonna ask you that.

SPEAKER_00 9:48

Yeah. Yeah. That's a big deal. So what's that? What are you saying? Currently, a lot of medical billing teams are just sitting on a backlog of denials and they just get to it when they can. But the great part about technology and software is that it can handle scale. So we can work the entire backlog of denials in a single day. So typically, providers will get paid 50% faster on their denied claims. And that also lowers their cost to collect, right? You don't have to have your team working as much or spending as much time doing this low value work where they can now spend their time on higher value work or time with patients.

SPEAKER_01 10:24

So at the moment, if I were using your software, I'd go through my normal process and I would engage your software only when I get a denial, not when I'm doing the initial filing.

SPEAKER_00 10:37

So our software is focused on the what we call the back end of revenue cycle. So after a claim has been submitted, how do you make sure this unpaid claim gets paid? There's definitely a lot of challenges on what we call the front end of like creating the form and submitting it. And there's a lot of technology focused on that area, but we focus on the back end.

SPEAKER_01 10:56

So how do you connect to the historical data when you're trying to match denial to information you have? Is that information put in by the biller? You're pulling it off their revenue system, or are you creating something as an in-between?

Raising $15M And VC Expectations

SPEAKER_00 11:12

Yeah, the whole goal is we never want to create more work for the users because they already have so much work. So we integrate with their systems, their EHR or practice management system, their clearinghouse, their insurance portals, and so on. So we're just pulling all the data from their existing systems. So have you raised venture capital for this initiative? Yeah, we've raised about 15 million. We're backed by Google, NVIDIA, Amazon, some great funds, a lot of VCs that focus in healthcare and finance as well.

SPEAKER_01 11:44

So given that you come from a bit of a venture background, which a lot of people don't come from, did you have these relationships previously? Or is there something in your awareness of how to package your business that makes it more attractive to get that 15 million?

SPEAKER_00 12:00

Yeah, a little bit of both. I think having worked in VC, I tend to know what gets VCs excited. So I can sort of tailor my pitch or how I'm thinking about the business or my answers during these meetings. And the venture fund I used to work with, they're great. They also invested in this company. So that was a great way to get off the ground.

SPEAKER_01 12:19

So as a would you say you're a healthcare IT company? Is that how you describe yourself?

SPEAKER_00 12:24

Yeah, exactly. Healthcare technology, SaaS, things like that.

SPEAKER_01 12:27

I'm interested in we have a lot of healthcare IT startup people that come from the provider side. And so they don't have that view of the venture capital and the acquirer that you would have, right? So maybe explain to our audience how a venture fund works, how it makes money, and what it's looking at as an investment would be really helpful coming from your perspective, particularly.

SPEAKER_00 12:49

VCs are in the game of very high risk, high reward investments. They're trying to find the next Facebook or Uber, and they'll sometimes make a hundred investments, hoping just one or two of them pay off. They know that these startups are going to be tackling huge, ambitious, messy problems, and most of them are not going to work out. I think probably the biggest shift from a traditional business is trying to position your startup as if this succeeds, this is going to be a hundred billion dollar company one day. This could be the largest company in the United States, not just a nice little small business where we can retire a little bit early, maybe. So that's probably the biggest difference, I'd say that when I try and answer questions from VCs, thinking as big and as boldly as possible would get them excited.

SPEAKER_01 13:40

I think one of the things that strikes me is when you talk to an individual or person who's coming at this for the first time, and they'll say, Why is everyone telling me I should have 10x multiple return? Because when you look at the fund averages, they're not averaging that. But to your point, the reason they say 10x for everybody is out of the hundred companies, one or two make it, and the rest do it for maybe singles, a couple doubles in there. But for the most part, the 10x is about how do I have that billion dollar opportunity when I have a hundred million dollar fund? Because that's what I need to get to with all these monies I'm deploying, right?

SPEAKER_00 14:23

Exactly. Yeah. For the venture fund, that one investment has to pay for all of the losers that go to zero, plus return the fund multiple times over because the investors in venture funds are looking for two to three times their money. If they were looking for 5%, they would go to the stock market.

SPEAKER_01 14:41

And what's interesting to me is in the healthcare IT companies that I work with, the joy and the danger. The joy is you can get to cash flow positive at a humble amount of money compared to their pharmaceutical company. The bad news is when you get to that cash flow positive, a lot of CEOs don't feel the pressure for that hockey stick earnings and throttle it back, and they end up not getting the results that they would get if they were doing a drug or a medical device. Because with those, you're putting in 50 to 100 million and then you're looking for the big return after. So you've got to swing for it to get the money to begin with. What advice do you have for people that are in that sort of situation when they get to the humble point?

SPEAKER_00 15:26

Yeah, it's tough. I mean, transparently, we're not at that point, right? In startup world, it's often that if you reach that profitability point, you're not being aggressive with growth enough. Yeah. You should be investing in your sales or your marketing so you can capture and win the market. Because if you try and grow conservatively and stably, someone else is going to capture the opportunity. So I mean, I think it's a great question, but maybe I'm not qualified to answer it.

SPEAKER_01 15:52

Well, I figured you on the VCN saw lots of these companies and you could see the difference between personalities from that perspective.

SPEAKER_00 16:00

Yeah, I mean, from that perspective, I think it all comes down to the founder because the reason they go into starting a company, it's a very challenging thing to do. So they tend to be so driven by the problem they're solving that just because they've reached this point of profitability, it doesn't mean their mission is complete or the job is finished. So I think it just comes down to, like you said, the personality of the founder.

Selling Into Healthcare Pain Points

SPEAKER_01 16:23

So when you look at IT companies and software companies, generally the big question is show me how you're going to get into the sales channel, right? So how do you get into the sales channel and how do you control that space? Is it that you're a first-to-market mover and that's why it's there, or has this been a gap that's been there for a while?

SPEAKER_00 16:42

Yeah, totally. I think selling in healthcare is one of the most challenging industries to sell into because it's really hard to get your doctor on the phone, even if you're a patient. So to do it and try and sell them something is way harder. I think what you mentioned is true is that this is a gap that's been here for decades, right? We know moving money and paying for healthcare has been broken for decades. And it's such a burning need for these healthcare providers because it's the lifeblood of their business. If they don't get paid, they're not in business. So this is usually one of the top things that keeps the doctor or the owner up at night is how do I make sure I have money for payroll next week? And so when you can sell to a pain point like that, they'll usually give you the time of day and hear you out a little bit.

SPEAKER_01 17:30

And I think for the audience, we've all had a new job where you don't get your first paycheck for a couple of weeks. Or in my case, I started in the defense industry. It took a month when I got out of school. And so that was rough. We're talking with doctors going 90 to 180 days, where actually they're funding the business.

SPEAKER_00 17:48

Absolutely, yeah, because the what we call the revenue cycle takes months, right? Insurance companies don't pay you out the same day. They often take weeks or or months to pay out. So when a doctor does your root canal today, they're not getting paid until February. Yeah, that's nuts.

Building Trust With Collaborative AI

SPEAKER_01 18:06

So, what challenges your organization facing as you see the future of healthcare in the next few years and how you're matching up to it?

SPEAKER_00 18:13

Yeah, I think it's moving so fast, especially with AI. I think we're seeing AI become relevant in almost every workflow, both administrative and clinical. A huge barrier to adoption right now is that healthcare providers don't fully understand it because it's moving so fast, right? They don't know what data is it using, is it safe, is it compliant, is it ethical, what happens if it messes up? So there's just a lot of hesitancy right now in the industry. But I think we're starting to see some use cases move a little bit faster than others. So that's a challenge for us is how do we make sure when we're dealing with something so sensitive and important, like finance for a business, how do we give them confidence to actually use it and build that trust?

SPEAKER_01 19:00

So during your journey, what has been a moment where you've had to adapt a shift that surprised you? And what was it and how did you handle it?

SPEAKER_00 19:07

Yeah, I think when I first set out on this business, we had this promise of like, you know, the AI is going to do the billing for you. And I thought that was great. Well, what I found was that was actually scaring them. Like I mentioned, they don't really trust AI very much to do billing. So a huge change we had to make was how do we position our product as a way that doesn't, it's not like a black box that you just say make and fix this claim. It's how do we give the existing medical billing staff AI tools so they work in collaboration with AI so that they remain in control, they have visibility and transparency into what's going on. And at the end of the day, they are the driver, not AI. And that was a huge shift we made that has landed a lot better given where we are in terms of AI adoption in the industry.

SPEAKER_01 19:57

So it's more set up as decision support than that sort of black box. I think too that unfortunately, there's a lot of companies when you sort of peel back the onion a little bit, they're more of a service company putting stuff through make or N8N and not really having any proprietary data or search engine activities. How do you handle that competitively? Because there has to be some people out there claiming to do things like you do and they're really not.

SPEAKER_00 20:26

I think it's kind of an interesting strategy to be able to sort of mock the functionality before you have it. And when we first started, we just said we provide medical billing just so we can get in the industry and learn. And it was 100% services. But I would say that it's actually easier for us to compete against those more services-based companies because it's still providing a lot of uh automation benefits versus what exists today. So, like those services companies are basically on par with the healthcare providers' billing team. They're suffering the same problems, they can't get on the phone with insurance companies fast enough. They're stuck manually drafting appeal letters. So it actually helps us when we go into a competitive sales deal against a more services-based company.

Hiring Billers Beside Engineers

SPEAKER_01 21:16

Well, your founder's story is fantastic. What challenges do you have finding people as you scale? I mean, you've got technical people, but you also have reimbursement people, I imagine, right?

SPEAKER_00 21:26

Yeah, it's kind of interesting because when you're building AI for such a specific domain, you have to get that domain knowledge somehow. So we actually have like a director of Revenue Cycle who's done medical billing to help us inform what we're building. We also have trained medical billers on staff that make sure we're building things the right way. They're verifying the AI is correct, and that helps us train the AI to make it better and smarter. But yeah, it's a huge challenge. So we have medical billers sitting next to engineers in our company.

SPEAKER_01 22:00

I would imagine actually for those folks, that's fun compared to manual pools of people that they've been in. So how do you keep current with all the changes? Because the other piece about billing, and maybe people don't know this, but at least annually there's changes going on, if not rulings during the year, right? So that's also a big surveillance effort on your part.

SPEAKER_00 22:20

Yeah, you're totally right. The AI is only as smart as the information it has. So we've put a lot of effort into making sure we have the most up-to-date policies, changes, guidelines, rules, all this stuff from all the government organizations and entities that define these rules. So it is a huge engineering challenge to scrape that data. Something that's also challenging is that every insurance company has their own policy, which is, you know, when you're a healthcare provider and you go in network with an insurance company, they'll have you, they'll send you the policy or they'll say, you know, you agree to the policy. But those policies change all the time and they're not published very often. And so it's a huge challenge to try and get that data set from insurance companies.

Fraud, Errors, And Compliance Reality

SPEAKER_01 23:06

I would think too that if I were worried about misbilling or perceived fraud, which is very often an error and not intent, your system is of great comfort to me because it's a big delegation for a physician. Practice to be able to say, okay, these people are getting my money.

SPEAKER_00 23:25

Fraud, waste, and abuse are massive concerns for the insurance companies. Because, like you're saying, it's very easy for providers to take advantage of this. So the insurance companies have huge departments that are very sophisticated. They're using machine learning to identify irregular patterns. And when I've spoken to them, they're also very cautious about using the F-word fraud. They often think it could just be errors or someone just doing something the wrong way without knowing. So they have a very high burden of proof before they say the F-word. But like you're saying, it's going to be a huge relief to providers when they know that the work they're billing for is being billed compliantly.

SPEAKER_01 24:06

Yeah. I did a project a few years back on artificial limbs. And I'll use hangar as an example, big manufacturer of artificial limbs, and then all these little mom pop places where you go to get them fitted. And there was something like 32% of the bills were denied, and 25% of their payments never get paid. And it all had to do with documentation. And then the other issue they had was the insurance program only giving certain reimbursements on certain limbs that didn't fit people. And so they'd get the leg, which is not, I mean, I'm going to say, you know, full-length leg was like$20,000,$25,000. They'd wear it, it would bother them because it didn't fit perfectly, because that's all they could get. And they threw it in the corner. And then by the time they came back in two to five years, whatever the next step in their journey was, they became obese. They ran a wheelchair. And so lots of impact because of that inability. Have you looked at that space at all?

SPEAKER_00 25:10

No, we actually haven't. We've been primarily in specialties like mental health, orthopedic surgery, physical therapy, but not involved so much in like medical devices or artificial limbs.

SPEAKER_01 25:23

Boy, orthopedics is like a Lego set of very much needed. So what's the biggest lesson you've learned in your journey thus far?

SPEAKER_00 25:31

I think in starting a company and trying to grow it, every day is a roller coaster. There's so many ups and downs. And the biggest learning I've had is you just have to keep going and keep your head up because you're going to be a punch in the face many times every day. And if you take everything personally or too hard, you're never going to make it. So the advice I would give is always try and look optimistically around the bright side. And you have to get through those troughs to get to the peaks.

SPEAKER_01 26:00

Do you have a group of people that you stay in touch with? I used to run an accelerator and an incubator. And I used to say CEO is the most lonely experience in the world because your employees don't want to hear you're running out of money, your board doesn't want to hear it, and definitely your family doesn't want to hear it. So who do you talk to? Right. Some of the benefit of these accelerators is you could talk to people that are non-competitive. And how do you keep sane and have people to talk to?

SPEAKER_00 26:25

That's definitely true. I'm fortunate that our investors, especially the ones that are more engaged and active, want to hear the good as well as the bad because they want to help. So they're often a really great sounding board. Or if I'm stressed or frustrated, I can always bounce ideas off them. And I also have a co-founder who's great. So it's someone you're in this together with. So we can often voice our frustrations with each other.

SPEAKER_01 26:49

So I think that's an interesting perspective you may have gotten from having been in venture capital, that it's a two-way conversation and the right investors are also there to help. I'm not sure that everyone would have that perspective. It's sort of like feeling like you're giving the teacher the wrong answer to some people and they kind of keep it to themselves and they don't have an outlet for that.

SPEAKER_00 27:14

Yeah, I think that's natural. And that's something I was doing at the beginning was I was sort of scared to say when things were not going well. But then I found like the more I did, the more they were able to help and like turn things around.

SPEAKER_01 27:27

Well, and then I think they also get to understand how you think. Absolutely. When a venture fund is trying to ask themselves the question if they trust to see part of it is knowing how they think and will behave. And so I think when you're sharing those wins and those losses, that's very helpful from that perspective. So what do you see as the biggest opportunity and or threat for you in the next few years as you grow?

SPEAKER_00 27:53

Biggest opportunity is that the market is so big, we are working at less than 0.01% of it. So there's a long way to go to try and fix this healthcare payments challenge. Biggest threat is this is a huge market. So there's going to be other startups and incumbents trying to tackle this problem. We know we're not the only ones. Huge threat could be that we're just too late and someone else gets there first. Interesting.

SPEAKER_01 28:20

What else would you like to share with the audience?

SPEAKER_00 28:24

I think we covered a lot of things. Yeah. How can they find you? Yeah, you can find us at ArrowHQ. That's our website. Well, thank you for being guests today. Really appreciate it. Thanks for having me.

SPEAKER_01 28:38

Thanks for tuning in to the Chalk Talk Gym podcast. For resources, show notes, and ways to get in touch, visit us at chalk talkgym.com.