Commure raises $70M at $7B valuation to deploy AI agents across the full healthcare administrative stack

Speaker 7: you Thank you guys so much for having me.

Speaker 1: Cheers. Have a good one. Goodbye. Up next, we have Tanay Tandon from Chemure.

Speaker 2: He's back. No. He's back. Raising 7,000,000,000 at a $700,000,000,000 valuation.

Speaker 1: Not too far from it. I'm sure he'll be there soon. Welcome to the show. How are you doing?

Speaker 8: How are

Speaker 6: you guys?

Speaker 9: Thanks for having me.

Speaker 1: Good good entrance, drinking casually. Yeah. Oh, dial.

Speaker 2: Oh, guys.

Speaker 1: Oh, hey.

Speaker 2: See you Good there.

Speaker 1: I'm just live. Anyway, welcome back to the show. Please reintroduce the company. Tell us the news. I want to hit the gong and hear all the greatest the latest and greatest.

Speaker 9: Awesome. I'm Tanay, CEO of Kemira. We just announced a raise of $70,000,000 at a $7,000,000,000 valuation with

Speaker 1: This guy hates dilution. He hates dilution. Only 1%

Speaker 9: with GC, Sequoia, Morgan Stanley, Kirkland Ellis.

Speaker 1: Yeah. How how do you get to this? Is this more of a strategic round? Did you give it a name? Is this this particular letter or was this more opportunistic and you have a particular goal in mind to take it to the next level? Like, what's on the horizon for the next year?

Speaker 9: Yeah. One, it's an extension. It's like, I think we called it officially a series e one or Sure. E two or something like that. Yeah. The goal, I mean, one, it was we didn't need the cash. We thought it would be a good time to market the company at a fair price for all the work that's been put in over the last eighteen months. And then on top of that, take some cash, put it on balance sheet to really accelerate R and D around some of our investments on Air, which is our language model powered EMR platform, Ambient and voice agents. Hire a group of forty, fifty elite engineers and just hit the pavement.

Speaker 1: There we go. Very cool. How much of the I mean, it sounds like you're already expanding outside of, like, revenue cycle management, like more back office workflows. I'd be interested to know the the shape of the business, some of the different products, how health care providers are actually integrating with you.

Speaker 9: Yeah. I mean, we see the problem as this trillion dollar administrative work tax on the American economy. You have 4 or 5,000,000,000,000 that you spend on health care, but the fact that 20% of that is spent on labor that pushes documents, submits claims, writes documentation is a travesty. And our belief is that language models can handle all of those tasks. So the core product lines, as you mentioned, is revenue cycle, which is an engine that takes claims, automates the submissions, appeals, denials, prior authorization process. Ambient documentation, which takes the workflow around actually writing notes that a provider might do with the patient and completely eliminates all the work tax around that. And then voice agents and back office agents, tools that automate scheduling, tools that automate the task of putting someone on a calendar, putting someone on a prior auth or appeal schedule, and just doing that with voice models. So those are the key areas and that's where we're going to continue to invest in more.

Speaker 2: Jordy? Every healthcare CEO historically will complain at different points about how slow moving adoption can be at times. Has that changed over the last two months? Are are different groups adopting, you know, new products and services much faster than they would have historically just because there are these pretty dramatic advancements?

Speaker 9: I think healthcare has been one of the areas alongside legal and I would say coding, like software engineering, where we have seen the fastest adoption of language models because it's just such a, you know, hammer on nail situation for for for the work that these providers are doing. And post COVID, I think we burnt our providers out. Most of these providers were working, you know, fifteen, twenty hour hour days and just not getting much sleep. Many of them wanted to leave the health system and go work in tech or finance or something easier. And language models were the gift that arrived at the right time to keep them in in the workforce that we need them in so much.

Speaker 1: Can you talk a little bit about invisible AI to I'm wondering how much of your product sort of like reveals itself to be AI powered to the end user, the customer, the person actually receiving health care? Because I think there's like maybe some sort of transition happening where members of a health care organization are using AI, seeing speed ups, but the actual end user, the customer, the patient might not even be aware that AI is involved at all.

Speaker 9: I think the beauty of language models is you can truly sell the outcome. There's like a big Twitter thought piece right now, but we live it in the sense that we sell the outcome of more revenue for a practice or a health system or better documentation for a practice or a health system. And the way to do that isn't necessarily brand and market yourself as an AI enabled this or that, it's just deliver the amazing result for a price that's a hell of a lot lower than the rest of the market. And I think for revenue cycle, for example, it's been an end to end service that's been provided with offshore labor in India or Bangladesh for twenty, thirty, forty years now. And we're taking that model and instead deploying agents on that same task and delivering a better product at a lower price.

Speaker 1: Are you already seeing evidence of, like, agent on agent conflict or collaboration, I guess? I'm imagining that, like, you know, a power revenue cycle management tool winds up sending me a bill or customer bill, and then they're open clause debating it. And, like, what does that future look like in your opinion?

Speaker 9: I I think there's the collaborative piece that you alluded to, which is super exciting where you see models literally coaching other models

Speaker 4: Sure.

Speaker 9: Creating better prompts, creating iterative versions of the same, know, task execution methodology, and we have a lot of investments in that. We've seen over overnight generation across hundreds of thousands of claims. The same model performs ten, twenty times better than it did when it started. And then there's the kind of combative models where you have insurance companies putting up their own nonsense models, trying to deny claims, and then our models are fighting those models. And it really will turn into, in some ways, a war of attrition. I think the final end state there is you have models talking to models, you eliminate the labor costs, and you take healthcare from this 15% cost to collect business and turn it into a Visa, MasterCard like business, where there's 3% interchange fees and it returns billions, if not trillions to the health system.

Speaker 1: Sure. Are you because of the maybe you can give a brief overview of like the structure of the health care system because I think people sometimes misunderstand how consolidated the insurance side is versus how diversified the provider side is. But then I'm interested to know, are you permanently in a lane or do you have business to do with all sides of the market in the limit?

Speaker 9: Yeah. I mean, first first of all, we are like a provider first and provider only company. I think the the provider is the only protagonist in our story and we think of ourselves at times as an arms dealer for the provider. Give them the tools to go nuke the payers and and really get their margin back.

Speaker 1: Yeah.

Speaker 9: In in in the context of, you know, the the the broader like payer ecosystem, I think one of the concerning trends is this, like like you mentioned, there's just sheer volume of consolidation. You have payers that are essentially monopolizing and dictating how much providers get paid for every little thing. Then on top of that, denying, denying, denying, which makes it way harder for a provider to earn a living. Compare that to the nineties where providers were making money hand over fist and living good lives and I think the quality of care in America was better back then too.

Speaker 1: Yeah. Are you is there is there a reason to be generally in favor of provider consolidation sort of paradoxically because the payer ecosystem is so consolidated that the providers can't push back at their current scale. And maybe some of the roll ups and mergers that we're seeing on the provider side could actually create sort of a strength that might actually benefit the end consumer.

Speaker 9: We see both sides of that coin. Mean, one, we're partnered with HCA, which is literally the largest health system in the country. It bills over a $100,000,000,000 in revenue a year. But on the flip side, we think AI and language models create this opportunity for more independent practices and more physicians starting their own businesses. Now the reason why I think both of those are interesting, if you have a tech layer that lives on top of both, that almost becomes the GPO or group negotiating organization that can lower or that can improve pricing and negotiate better rates against payers, kind of like, you know, like the flip side of the whole ramp vendor management tool or one of these other software spend management tools where you consolidate and add price transparency and then you return margin back to the entity that used the tool.

Speaker 1: Yes. Yes, that makes sense. Are you seeing any evidence of an uptick in individual practices in or is it too soon? I mean, we're seeing like a lot of solo entrepreneurs. Every entrepreneur wants to like build the $1,000,000,000, one person tech company, but it's usually like a vibe coded piece Pretty of soon.

Speaker 2: Will see the one doctor, doctor, $1,000,000,000 hospital.

Speaker 1: Hey, maybe if they save the right person's life, you know, willingness to pay.

Speaker 9: I I think the the thing that we are seeing for sure is the practices that have been independent are becoming higher margin and becoming more profitable when they adopt AI tools. Interesting. And that's I think the first step and a necessary precursor to the creation of more independent practice because one, you're going to have them begin to invest in other practices or potentially roll up practices. You're also probably going to see this concept of the AI first practice, like a truly online behavioral health practice that uses LLMs for everything except for the care. You're definitely seeing this in the pharmacy world where there was like the recent New York Times article about the GLP-one business that it scaled to a couple 100,000,000 in run rate.

Speaker 6: Yeah.

Speaker 9: And I think you're going to see more and more of that across the ecosystem because of language models.

Speaker 1: Interesting. Well, congratulations on the new round. Thank you so much for joining the show. Jordan, anything else?

Speaker 2: Great to see you.

Speaker 1: You good? Thank you. Congrats. Have a great rest of day. To you too.