This week’s newsletter is brought to you by: Alumni Ventures
If you’re curious about investing in health tech, Alumni Ventures Healthtech Syndicate lets individuals invest in startups backed by top firms like a16z, Lightspeed, and Khosla Ventures.
It’s free and easy to join. Members get early access to curated deal flow including full diligence materials, with the option to invest starting at $10K.
Alumni Ventures shares everything you need to make confident decisions — and connects you to a community of like-minded investors.
HTN is not affiliated with Alumni Ventures and does not provide investment advice.
VOICE AI
What’s going on in the agentic voice AI market?
This week, we saw large funding rounds announced within a few days from two companies competing to become the leading end-to-end patient-facing agentic voice AI platform for care delivery organizations.
On Tuesday, Prosper AI announced $30 million in funding from a16z and others. Prosper reported it has grown revenue 5x in the last six months. On Wednesday, Assort Health announced $120 million from Menlo Ventures and others, at a $1.2 billion valuation. Assort also reported 20x revenue growth over the last 15 months, seemingly implying it is currently at around ~$60 million in ARR (based on previous revenue estimates).
It’s fascinating to peruse the press releases and see how both describe their ambitions to become the leading agentic AI platform in the market. Their customer quotes are oddly similar in how they describe the differentiated platforms for each company:

I know these are startup funding announcement and so anything tends to go with these announcements, but the quotes are a bit curious, right? Here you have two customers, claiming to have done a comprehensive evaluation of the market before determining that their vendor is the only platform offering that meets their needs. M
eanwhile, if you peruse the websites for Assort and Prosper (not to mention a number of other vendors in this market) the platform capabilities sound the exact same: agents for scheduling, eligibility, billing, and so on. Not to mention, the results weren’t exactly encouraging when I tried testing one of Assort’s agents in the wild after Assort last raised funding in the fall. Either these customers are evaluating their platform needs very differently, these vendors are operating in different segments of the market, or these evaluations aren’t quite as robust as indicated. None of the explanations seem to inspire confidence that these are mature, differentiated platforms winning on their merits.
Perhaps none of this actually matters, as these companies are reporting massive growth via their preferred up-and-to-the-right metric of choice in the charts below. It seems illustrative of the general point that each company is picking a different metric to highlight its impressive growth trajectory:

Source: HTN; company press releases
The underlying rationale driving the growth and investor interest in these companies seems pretty straightforward — medical practices are in dire need of administrative support, and agentic AI is pretty effective at handling many of the tasks, particularly when these practices are struggling to staff people in administrative roles. The numerous case studies available on agentic voice AI startup websites point to this, along with the growth in the charts above.
For instance, check out this Prosper AI customer, an OB/GYN group in the Northeast that saw a $100k EBITDA improvement by getting more efficient at scheduling, apparently in a matter of weeks. Or this Prosper customer that reports $1.1m in annual labor savings. Or this Assort customer that is seeing a $1.3m revenue increase, with $400k in labor costs avoided. Or this one that reports a $1m revenue increase. Those results are all quite impressive. It’s not hard to see why these companies are growing so quickly at the moment.
All of this activity prompted a good conversation in Slack this week about the size and durability of this agentic voice AI market. There clearly is an opportunity, particularly within the PE-backed specialty practice subsegment of the market, to help improve financials using AI voice agents in the short term. We can see the companies racing after the opportunity here. What is much less clear to me is how many companies this market can support and what actually wins over the longer term. Right now, differentiation seems to be primarily in GTM, getting in front of customers and grabbing them by demonstrating financial results in weeks. While these companies are claiming to be platforms that are winning rigorous head-to-head evals against the market as platform plays, that seems quite unlikely to me. What seems much more likely is that this is a land grab of practices exiting the pre-AI dark ages, with vendors winning contracts by highlighting the financial upside for those practices. It all raises a question of how durable this opportunity is as these capabilities become table stakes.
It can’t persist forever that there’s essentially a free pot of money for medical groups to capture by actually answering their phones. At some point in the relatively near future, it seems inevitable that this inflationary environment will subside, which will invite practices to evaluate more closely what they are spending dollars on. I think its worth thinking about what market dynamics here look like when that happens.
I am not particularly bullish that this specific market will support multiple billion+ dollar outcomes, which these venture rounds seem to imply. Companies like Assort appear to be gearing up to roll up the market, and the best case valuation outcome seems to be winning the rollup race and timing an acquisition right by a larger organization experiencing FOMO.
In Partnership with Nashville Health Care Council
From Sep 13-15, you can find us out at Nashville Healthcare Sessions, a substantive conference limited to ~1,500 attendees across incumbents, innovators, and investors discussing AI, VBC, health policy, and more.
Register with code HTNSESSIONS for $100 off.
CHART OF THE WEEK
The counterintuitive results of mapping physician supply versus average wait times
I’m a sucker for a good musing on physician shortage data, and this Substack post by Joel Selanikio is no exception. It makes the point that Boston has the longest average wait for a physician in the US (compared to other large metros below), despite Massachusetts having more physicians per capita than any other state in the US. That point shows up visually in the chart below. Intuitively, I’d think the opposite would be true — more docs per capita should be correlated with shorter wait times, no? It invites lots of questions for me about supply / demand and how we measure these sorts of things.

Other Top Headlines
Cityblock is reportedly acquiring Homeward Health, gaining Homeward’s in-network Medicare Advantage contracts in Michigan with BCBS Michigan and Aetna. Homeward launched in early 2022 with $20 million in funding from General Catalyst, which is also an investor in Cityblock, and a narrative about how it was going to rethink rural care delivery. In August 2022, Homeward raised another $50 million at a $345 million valuation, as it gained early traction in Michigan, in what ultimately appears to serve as another tough reminder of how frothy that investment environment was. Homeward provides Cityblock with some interesting diversification opportunities to build upon in Medicare Advantage and rural care specifically.
UpDoc, a clinical AI startup, announced that its LLM received FDA clearance, along with $18 million in Seed funding. The WSJ Pro coverage of the clearance discussed UpDoc as a “concierge doctor,” although it appears the 510(k) clearance the company received is much more narrowly scoped to titrating insulin doses than what I’d think of as an AI doctor. Following the approval, the Cleveland Clinic will now initiate a 6 - 12 month pilot to see what the impact is on patient outcomes, which I thought was the point of an FDA approval. It raises lots of interesting questions about the right regulatory framework for AI in clinical care, and whether AI should be treated more like a medical device (as here) or more like a clinician practicing medicine. The regulatory approach being tested in Utah, which aligns to treating clinical AI much more like the practice of medicine, is one that conceptually makes much more sense to me.
Early analysis ACA rate filings for 2027 by Modern Healthcare indicates that payers are planning for another year of high premium rate increases, with a number of payers asking for 10%+ increases.
Elevance Health paid CMS $342 million as part of resolving the risk adjustment issues that led to CMS implementing sanctions that would have suspended new enrollments in Elevance MA plans later this year.
A NUMBER TO PONDER
$9 trillion
The latest national health expenditure data published in Health Affairs this week suggests that it will hit $9 trillion in 2034, representing 20.6% of GDP. Those numbers are up from a reported $5.3 trillion and 18% of GDP in 2024.
Funding Announcements
Alan, the French insurtech platform, raised ~$555 million, valuing the business at ~$6.3 billion. Prosus led the round. Alan’s CEO shared on LinkedIn that it generates ~$925 million in revenue, is growing at 53%+ annually, and is “profitable in France.” It seems worth noting that Alan is also live in Belgium, Spain, and Canada, which would seem to imply that the business is not actually profitable. We’ve seen the challenges that insuretech startups have had getting to scale profitability in the US over the last few years, so will be interesting to keep an eye on how Alan grows into it 6x revenue multiple.
Trase, an agentic operating system for regulated industries, raised $107 million in Seed funding. ARCH Venture Partners led the round. Trase’s press release highlights its work underway at Duke Health, where it has automated 5,000+ monthly faxes for the cardiology department, unlocking $285k in annual staff capacity for the department. Seems like a pretty logical AI deployment use case to me. Venture studio Red Cell Partners launched Trase with $10.5 million in pre-seed funding last fall.
Cadence, a remote patient monitoring platform, raised $100 million. Spark Capital led the round, along with an impressive list of other investors, including Thrive, General Catalyst, Coatue, B Capital, Memorial Hermann, and Duke Health. Cadence’s narrative in the press release is much broader than RPM, as it articulates a vision around an AI-powered remote care model for health systems like Duke. Cadence shared its ambition to scale up to 10 million patients, from its current state of 100,000 patients working with 20 health systems today. It’s a logical expansion for Cadence, given both the interest in AI care delivery models and the increasing scrutiny on fee-for-service RPM billing practices. Mario Aguilar’s coverage in Stat included a nice example of how this will be a meaningful shift for some of Cadence’s existing health system customers, highlighting how Rush will need to rethink billing while also running it through an AI compliance review process.
Hera, a care coordination platform for original Medicare beneficiaries, raised $27 million. Bain Capital Ventures led the round. Hera is providing care coordination services to over 1,000 seniors since it launched last year, with plans to expand into 25 states this year. It appears to be leveraging Chronic Care Management (CCM) codes to provide these services to original Medicare beneficiaries, leveraging 1099’d RNs and Social Workers to bill incident to providers. It’s one of a handful of plays we’ve seen recently in this market, and we’ll be keeping an eye on how this market unfolds over the coming years.
Upside, a platform addressing housing instability, raised $20 million. Aquiline led the round. Upside is live in 10 states focused on the Medicaid population, and has contracts with 20 payers. The engagement numbers are quite impressive: Upside reports connecting with 90% of eligible members, enrolling 85% of those it connects with, and stabilizing housing for 50% of all eligible members. While we generally talk about the headwinds facing the Medicaid market, it certainly appears that there are some pockets like this that are seeing momentum. Every time I think about payers addressing housing needs, I go back to this excellent Bloomberg article on UHG’s MyConnections program from back in 2019, which does a really nice job outlining both the societal opportunity at hand and the finances of these sorts of approaches.
What I’m Reading
I am quite curious about the concept of “lasering” in the stop loss market, which is the idea that carriers can essentially carve out high-risk individuals from a group policy by applying a significantly higher deductible to that one individual. It is being used to navigate high costs around oncology, specialty drugs, and complicated surgeries. Marsh and McLennan recently highlighted how it helped a 600-person company save ~$1 million by recommending a “laser” for a single employee who was potentially going to have a high cost procedure. Marsh did an analysis, determined they didn’t expect the procedure would happen in the next year, and negotiated the laser cost down to $700k versus $865k. I have lots of questions about the ripple effects of this if this approach continues to emerge.
Michael Stratton and the Health Data Atlas crew shared a good explainer on the growth in freestanding emergency departments, and the implications of their growth in a handful of states. Texas is the leading state here, apparently with roughly 25% of ED visits treated in freestanding EDs. It highlights how HCA, Atrium Health, and Nutex Health are some of the bigger players in this market. Regular HTN readers will recall that Nutex Health has been causing a stir in Idaho over its aggressive billing strategy centered around remaining out-of-network and leveraging the No Surprises Act.
Speaking of No Surprises Act shenanigans, Elevance shared data looking at the 25 most common planned surgical and endoscopic procedures in Elevance’s IDR data. The differences in payments between the IDR process and FFS Medicare is staggering, with a median payment 67.4x higher and the mean payment 144.8x higher. While I know both providers and payers feel strongly about their respective arguments around the NSA, these sorts of results continue to make zero sense to me.
Featured Jobs
Manager, Transformation, at Aledade, a primary care VBC enablement platform.
$120K - $150K | Remote
Hopscotch Primary Care, delivering tech-enabled primary care to rural communities, is hiring for the following roles:
Manager, Analytics & Strategy $120K - $140K | Hybrid (Chicago)
Security Operations Engineer $90K - $100K | Hybrid (Chicago)
Senior Data Engineer $150K - $190K | Remote
Strategy & Operations Manager at Rula, a digital mental health care company.
$136K - $152K | Remote
Contact us to feature roles in our newsletter.
Want to comment or share feedback?
If this newsletter was forwarded to you, subscribe here or see more from Health Tech Nerds.





