Kristin Myers: More than 4 in 10 American hospitals are now losing money on their operations, and some of that money isn’t spent on care. It’s actually left on the table. That’s rebates nobody claimed, contract prices nobody enforced, insurance payments that came up short. My guests today are building AI agents that go find that money and then help get it back. And last week they raised $20 million to do it. But first, let’s take a look at the company. The road from an idea to the opening bell starts right now.
America’s hospitals have a money problem. And it’s not just about what they spend. It’s money they may already be owed but never collect. Unenforced pricing terms, unclaimed rebates and insurance underpayments can all eat away at hospital margins, and that’s the problem Sumit Kadakia and Venkat Mocherla are trying to solve.
They co-founded Midstream Health, using AI agents as a second financial brain to find and recover money across supply chains, pharmacy and managed care. And now it’s finding scale. Midstream says its technology now reaches more than 250 hospitals, including some of the country’s major health systems.
Now, Midstream has raised a $20 million Series A, backed by [unclear: “altruist”], Andreessen Horowitz, Foundry and CommonSpirit Ventures. But can Midstream prove AI can do more than help doctors and become the financial brain, helping keep America’s hospitals healthy? From Startup to Stock Exchange, Midstream Health co-founders Sumit Kadakia and Venkat Mocherla are next.
Joining us now is Sumit, who is the co-founder and CEO of Midstream Health, alongside his co-founder Venkat Mocherla, who is the president. Sumit, Venkat, thank you guys so much for joining us today. So, Sumit, I want to start with you and dive a little bit into your background. Now, your dad was a rural doctor in West Virginia, and the hospitals that he worked at have since closed. How did that experience lead to Midstream?
Sumit Kadakia: [unclear: “I think it and I on it’s”] a pleasure to be here. You know, really this all did start for me with these rural hospitals closing in West Virginia. When I say rural, I mean really one-stoplight-town types of places where care is delivered out of trailers, mobile homes, double-wide trailers, single and dual physician hospitals.
And my dad moved around a ton across the state when we were growing up, and I spent a lot of time, actually, in those hospitals and working in his practice. And in particular, there were two hospitals, Kristin, that closed that really impacted the way that I think about this problem. One is Elkins Memorial in Randolph County in West Virginia. And the second was Ohio Valley Medical Center in Wheeling.
The Ohio Valley Medical Center one was particularly impactful because 11,000 people ended up losing their jobs simply because the hospital couldn’t figure out a way to stay financially afloat. So given that I grew up in those communities, I know people that worked in those hospitals and how they were impacted. And recognizing this is a bigger problem than just those two systems, but it’s especially acute in rural America, made it all the more worthwhile and meaningful to say, let’s go figure out if there’s a way to leverage technology and leverage AI to be able to solve some of these problems and to help these hospitals out.
Kristin Myers: Now, Venkat, you know a little bit about tech and AI in this space because you watched hospitals adopt AI at [unclear: “Juventus”] and then at Andreessen Horowitz. So what gap were you seeing that nobody was filling?
Venkat Mocherla: Yeah. And thank you again for having us. For me, I spent the last 15 years at the intersection of building new companies and technology and health care. And for me, I think you had those stats up there. It was just fundamentally not acceptable in one of the wealthiest countries in the world to have not just 110 hospitals closed, more hospitals closed in the last decade than [unclear: “the last 100 combined”], but another 700 at the risk of collapsing.
And having grown up not only as an intern at my grandfather’s clinic, but around technology in Silicon Valley, and [unclear: “you”] firsthand how AI and agents can actually help hospitals recover dollars and make a difference of keeping the doors open versus closed.
Kristin Myers: So I’m going to throw this question at you, Sumit, but Venkat, if you also have a response, feel free to jump in. I think when most people hear that the health care industry is losing money, they’re probably thinking, yeah, right, because we hear so much out there about how much money is being made in the health care industry. So where is the money that hospitals are losing? In what areas, and why isn’t anyone catching any of those losses?
Sumit Kadakia: Yeah, Kristin. I mean, the unfortunate truth is that both things are valid at the same time. So if we zoom out, hospitals receive over $1 trillion a year in health care payments. And despite the fact that they receive $1 trillion a year, they operate on razor-thin margins. So that’s why there’s a lot of money that goes into the system.
And I believe that we need to make sure the hospitals are armed with the right tools and the right technology to be able to find those places where dollars are being leaked, where they’re not able to go deliver care to the people that need it. Specifically, we look at contracts across the ecosystem because an average hospital has anywhere from 3,000 to 6,000 contracts.
They have more SKUs than Walmart and Target combined that they are bringing into the system. They have highly complex labor forces, and so their ability to manage all of this complexity that insurers put upon them, that vendors put upon them, that pharma manufacturers put upon them makes it really difficult for them to be able to ensure that they are capturing just basic stuff like, hey, am I getting the right discount on this pharmaceutical drug that I purchased? Am I making sure that this payer policy update that just hit from my local Blues plan actually is reflected in the way that I operate correctly?
So it’s a lot of those small things that oftentimes don’t get picked up and don’t get integrated well into operations that Midstream focuses on, and that we help hospitals manage and recover.
Kristin Myers: Now, Venkat, and as everyone can see on the screen, we have something from one of your posts on LinkedIn where you guys essentially called AI a deflationary force. So I want to dive a little bit into that, because you’re calling it a second financial brain. So what are these AI agents actually doing on any given day?
Venkat Mocherla: Yeah, great question. Look, having been inside Fortune 500 companies where they have a ton of resources into data analytics infrastructure, you’ve got to remember our partners are mission-first organizations with razor-thin margins. So oftentimes, someone mentioned contracts, spend data, and payment data, all this data sometimes in different systems of record, and it’s being stitched together on Excel.
And so the first thing we do is have a single source of truth that acts like a second financial brain. The other challenge is, okay, how do you execute on that knowledge? And so now in this agentic era, you have these domain-specific agents that can go after supply chain, pharmacy, payer. And these very complicated financial decisions that require too much cognitive load can now be automated end-to-end and help these teams have, again, the difference between keeping the doors open or closed. Being in the red versus the black.
Kristin Myers: All right. So, Sumit, let’s talk a little bit about that AI, because a lot of the health care AI money has gone into billing fights with insurers. So I guess it’s not just everyday folks that find themselves arguing with insurance companies. I guess the hospitals are as well. I’m curious to know, why did you guys want to start there, with those fights with the insurers, with those costs, instead of maybe other places that AI might be able to improve billing efficiency?
Sumit Kadakia: Yeah, actually, I would say that we’re starting on the opposite side to correct that. So we’re starting on the cost structure of a hospital, not on the picking-fights side with the insurers. We are making sure that hospitals [unclear: “are just really basic things”] on day one. Are they paying the right prices for the products that they’re purchasing? Are they getting the right rebates for them? Are they negotiating appropriately, effectively and strategically for those contracts?
You’d be surprised, Kristin, that a single hospital can lose over $300,000 simply because they don’t have the tools or the technology to ensure that they are paying the correct price for a cardiovascular product that they are using in their system. So it’s a lot of, again, these things that are small and basic that ultimately help the hospital figure out what its actual operating margin is. And that’s really what our core true north is, what our core value is. Revenue is a piece of this.
But Venkat and I both believe, in order to make health system cost actually go down, in order to create that deflationary economy, we’ve got to make sure they’re picking up the nickels and the dollars that they’re currently just losing and leaving on the floor.
Kristin Myers: And why was it better to start with costs instead of those claims, and instead of spending so much time in those arguments with those insurance companies?
Sumit Kadakia: Yeah, it’s because those arguments with those insurance companies are an arms race that will only lead us to the same outcome that we’ve already seen for the past 20 years. Look, I think hospitals are tired of that game. Insurers are tired of that game. And most importantly, the American people are tired of that game.
So Venkat and I just sat down and we said, hey, we need to think about things differently when we as a country already spend more as a percentage of GDP on health care than everywhere else. CMS payments are not going to go up. So let’s go focus on how we use AI to actually make care more efficient and start there.
Kristin Myers: So, Venkat, Midstream has said that it’s found more than 1,500 recovery opportunities this year. So what does a typical find look like?
Venkat Mocherla: Yeah. I think what Sumit mentioned, the little details really matter. A missed form can oftentimes return a six-figure rebate payment that hasn’t been filed in time. I think what you’ll find inside these hospitals and health systems is there are complex financial decisions being made every day that require a ton of analytical analysis and timely submission. And frankly, the manpower and the analytic infrastructure is not there.
And so now what agents can do, magically, is not only analyze the data but actually execute on that, and actually find that rebate payment or make sure the right contract has been accessed for the right pricing that Sumit mentioned.
I’ll just sort of say, I think one curve that we keep looking at is, if you look at the cost of TVs in this country and compare that to the cost of hospital care and health care, it’s the exact opposite curve. And I think for the first time, as Sumit mentioned, we actually have a shot at changing the trajectory of health care costs. And I think that for us is the most exciting piece of all of this.
Kristin Myers: Now, Waystar makes the software many hospitals and doctors use to get paid, such as claims, denials and patient bills. It went public in June 2024, and today it’s valued at about $6.7 billion, and AI is now at the center of the pitch. Waystar says AI tools drove up about 40% of its new bookings in the first quarter, but the stock has struggled, well below its 2025 highs, as some investors worry that newer AI companies could actually disrupt health care software.
So here’s the contrast. Waystar lives on the revenue side, the claims fight with insurers. Midstream is betting the bigger opening is on the other side, the cost side: supplies, drugs and the contracts behind them. We’re back now with Sumit and Venkat, co-founders of Midstream Health. So, Sumit, I want to come to you here. We mentioned at the very start of the program that more than 4 in 10 hospitals are losing money on operations. You also mentioned this as well. How much of that is a cost problem versus a revenue problem, since the cost side is what you guys are focused on?
Sumit Kadakia: Kristin, it’s a really good question. It really depends on a market-by-market basis. There are going to be some hospitals, particularly rural hospitals, that are underpaid relative to what they should be making versus their more established counterparts that are urban. So in some cases it is actually a revenue problem. But our perspective is generally, across the board, this whole concept of thinking about things in terms of revenue versus cost is really outdated.
Venkat and I often make this joke that our local Chinese restaurant down the street has more information about their own margin and how their costs impact what they should be charging consumers when they come into the restaurant. Hospitals don’t work like that, which is really troubling. Hospitals either think about things on the revenue side, or they think about things on the cost side, instead of really connecting the dots and working with their patients to say, hey, here’s how much this actually costs to go deliver a colonoscopy and what margin we earn on that.
That is the way that hospitals need to start to think about things in the future. And that’s why our strategy, to your point, is pretty simple. Let’s start on the cost side. Step one. Step two, let’s help hospitals actually figure out what their margin looks like on a service line basis. And then step three, with that, we’re in a much stronger position to actually have real conversations with insurance companies about how care should be delivered, how costs should be controlled.
Kristin Myers: I want to ask you guys a little bit about Midstream and the expansion plans. But before I do, I want to throw this out. You guys have had a couple of graphics on the screen that we’ve shared, where you’ve pretty much talked about this zero-waste health care system, but how it’s also going to improve costs for end customers. I’m curious if you could speak a little bit about that, about how essentially recouping some of these costs can actually lower health care costs for a person who is coming into a hospital or seeing a medical provider.
Sumit Kadakia: Yeah, Venkat had mentioned this earlier, Kristin. But all of the institutions that we are working with are nonprofit and mission-oriented. So they try to take every single dollar that they can and put that into places like charity care, where patients that are underinsured or currently experiencing medical debt have the ability to go access care in a country where 1 in 4 oncology patients right now are faced with the risk of medical bankruptcy.
We believe that putting these dollars back into a health system’s bank account so that they can fund those programs is really, really important. So that’s the way that we see the direct linkage: if we give a hospital an extra [unclear: “five, $1,000 million”], just think about all of the additional care that they can provide with that annual additional margin that they have access to.
Kristin Myers: Now, Venkat, as I mentioned, I want to talk a little bit about the partnerships, about your investors. So CommonSpirit is both a customer and an investor. How did that come together, and what did that relationship ultimately prove for Midstream?
Venkat Mocherla: Yeah. Great question. And, you know, we wouldn’t be here without our partners. Every single one of them. And CommonSpirit is a great example. A mission-first institution that’s always been on the cutting edge of innovation. At Andreessen Horowitz, when I was a partner, I observed firsthand how their strategic innovation team and their leadership team, run by their CEO, Wright Lassiter, were always not only caring about the mission first, but also wanted to be a pioneer in innovating.
And I think what’s great about CommonSpirit is they’re in more than 25 states. They serve nearly 1 in 6 Americans. And for them to take a leap of faith on a company at that stage and at that time speaks volumes. Now we’ve returned the value in incredible ways. It’s been an amazing partnership, and we’re expanding on that.
Kristin Myers: And, Sumit, I want to ask about another partnership. Fairview and Keck Medicine of USC also just signed on. So where does the $20 million go when it comes to Midstream Health?
Sumit Kadakia: The $20 million for us helps fund expansion from focusing on med devices, which is where we built the initial AI engine and platform, into areas like pharmacy and managed care and purchased services. So for us, it’s all about expanding the breadth and reach, Kristin. Just with the initial runway that we’ve had to date, we already have systems that spend over $100 billion a year.
And so for us, that capital helps us continue to deliver services to those partners, but also to expand that and fuel this pretty tremendous growth that we’ve seen and the pull that we’re seeing from systems to go deploy Midstream in more and more areas and in more hospitals across the country.
Kristin Myers: So, Venkat, this is a question I ask a lot of founders on this program, particularly ones that are using AI, which is going to be around trust. Hospital finance teams tend to guard their data fairly closely. So how are you getting them to really trust these AI agents to act on that?
Venkat Mocherla: Yeah, it’s a fantastic question. I think I would sort of [unclear: “put an end in there”]. I think this game is entirely about trust and intent of that data. I think there’s two things that we’ve done. One is creating a source of truth for them, which is really meaningful. I think that oftentimes, knowing how to solve a problem, first you’ve got to know the problem itself, and if you’re operating off of a lack of a source of truth, that’s really difficult.
But the second thing I’ll sort of say is our company’s entire mission is to create a zero-waste health care economy and maximize savings for our health systems. It is not to monetize this data for selling, for example, to other parts of the industry, which a lot of, frankly, startups in the health system sector have done. And I think that’s exactly the wrong move. And so we are here for only one mission. It’s to help advance savings for our health systems, so ultimately they can do more for their patients.
Kristin Myers: What are you finding as you guys are expanding, that there is this sort of uphill climb or battle that you’ve had to really combat when it comes to the moment you start saying, “Actually, we have AI agents that are going to help you do this”? It’s a question that we’ve had some founders say yes, they’ve actually had to address in some industries, particularly finance. I’m wondering if you’re finding that same sort of reticence in the health care space.
Sumit Kadakia: Absolutely, Kristin, and I think everyone is trying to understand what AI is going to mean for the way that they work and for the viability of the organizations and for their jobs. And so we actually believe that this is a big responsibility for us to be the steward of AI, specifically in these parts of the health system that haven’t received a lot of attention from Silicon Valley.
Silicon Valley has been focused on rev cycle or on other parts of the actual health care financial stack. But these core areas, where a lot of hospital administrators work today, have never really had the experience of working with a technology company like Midstream.
And so we’re pretty intentional in terms of the way that we address AI, the way that we talk about it, and actually even simple things like we host hackathons with as many of our partners as we can and receive just really tremendous feedback from that. Because the real truth is that there’s a pretty significant lack of available labor to do this stuff. So this is not a jobs replacement type of initiative that we’re pushing, but much more of one of, hey, how do we just augment the people that are there to carry out these really complex tasks that hospitals haven’t been able to invest in before?
And as soon as we spend some time with them and go through that with them, they go from being initially skeptical to truly embracing what Midstream is able to do. And they get excited by it.
Kristin Myers: So we have about 30 seconds left, and I want to ask this to both of you. You guys aren’t here with me on the floor of the stock exchange, but everyone gets asked this question no matter what. I’ve got the bell right behind me. When you picture a Midstream IPO, what does that look like, and when do you think it’s going to happen?
Sumit Kadakia: Yeah. Well, we are growing like crazy, so definitely things like that are not out of the picture. But I’ll say, look, when we picture a Midstream IPO, Venkat and I have actually had this conversation, not only do we want, of course, our teammates at Midstream behind the bell, but we really want our health systems there as well. Again, this is a group effort that we’re all pushing to make sure that we’re able to help reduce costs in health care and drive the industry forward into a place that it should be.
So we view this as a really strong partnership with each of the systems that we have. And nothing would be cooler than to have our systems with us were that to happen.
Kristin Myers: All right. Thank you so much, Sumit and Venkat Mocherla, co-founders of Midstream Health.
