Sponsored by: UnityAI
We sat down with UnityAI’s founding team, a trio of experienced nerds who traded careers at HCA, a health system well known for operational excellence, to build agentic AI for smaller practices.
You can watch or read the discussion here, where we get into the following:
Why you should focus on data and process discipline before implementing AI
How coordinated agents can manage complex workflows
Where AI ROI is showing up
What resonated with investors in UnityAI's $8.5M raise
How UnityAI plans to use the funding to bring its product and roadmap to market
If you're interested in sponsoring the newsletter, let us know!
👋 Welcome to my Sunday recap of healthcare news. This week, I briefly considered pivoting and rebranding HTN to become NewNerd AI and going all in on AI compute. Alas, there was just too much going on in healthcare news this week, and so here I am once again sending you 3,000 words musing about ACCESS rates, ACA enrollment data, and the legitimization of the peptide gold rush.
- Kevin
ACCESS MODEL
ACCESS model participation and the staring contest between CMMI and VC-backed digital health… insurgents? incumbents?
CMS released the initial list of ~150 ACCESS participants on Monday, ahead of the program start date on July 5th. It also extended the deadline for companies to apply to participate to mid-May, seemingly indicating that CMMI would like to see more participation in the program. The initial list of participants is a healthy mix of medical groups, digital health companies, primary care enablers, D2C consumer models, and more.
But for the most part, it appears that the venture-backed digital health community chose not to apply to ACCESS after seeing the rates. Let’s go back to a slide from the model launch a few months ago, which highlighted 68 companies that had expressed initial interest in the model. I’ve gone through and highlighted whether that initial list applied or not (green = yes; red = no):
Source: CMS slide, HTN highlighter
That is a lot of red. By my read of the slide and announcement, of the 68 companies highlighted as having expressed initial interest, only 22 applied to the program. Almost 70% of the companies that expressed interest ultimately opted not to apply.
That outcome seems indicative of a staring contest occurring between CMMI and the digital health industry. CMS leadership has repeatedly made it clear that these rates are workable. In Mario Aguilar’s good write-up in Stat this week, CMMI Director Abe Sutton expressed his confidence in the design of the ACCESS Model. While CMMI appears confident in the model, the digital health industry appears equally as confident that the rate is unworkable, so much so that they chose not even to apply.
Take, for example, Sword Health. Sword has articulated a big vision for transforming healthcare via AI and its Sword Intelligence platform. It has also demonstrated interest in working with governments to do so — recall its partnership with Greece to transform the front door to healthcare there. As a private, high-growth business with a massive vision and corresponding valuation, conceptually, it is surprising to see it not participating in this program. I would imagine if there were any path to breakeven, or even to operate in ACCESS at a slight loss, a company like Sword would be jumping at that opportunity. So it seems like a telling example of just how far off these companies and CMMI appear to be here on the topic of rates.
CMS leadership has repeatedly indicated that ACCESS will be a program only insurgent models can succeed in, and that those insurgents will need to build a financial model from the ground up to succeed. Conversely, incumbents that attempt to pare back elements of their existing models will struggle. This is all part and parcel of launching a deflationary model. And certainly, it appears that we’re seeing indicators of that here. But at the same time, companies like Sword and others on the slide above are hardly what I think most folks would call “incumbents,” which is part of the confusion here. In many ways, this feels like CMMI attempting to cajole the industry to do better.
Ultimately, this disagreement seems answerable — there are dueling P&L models underpinning each of the CMMI and industry points of view. The question is, whose P&L is right? The fact that CMMI is moving forward without most of the digital health companies that initially expressed interest suggests CMMI remains confident that a group of companies involved will be able to make it work. The fact that the industry has looked and chosen not to participate indicates that digital health companies modeling the costs of AI-enabled care delivery cannot make the math work. The success of this model hinges on CMMI getting these numbers right.
As an aside, all the consternation over the financials here seems like a challenging data point for the “AI will drive the marginal cost of care delivery to zero” argument.
When looking for a path forward here, I’m particularly interested in keeping an eye on the consumer-grade device companies participating — Whoop and Withings specifically. It’ll be fun to chat with Withing’s Patrick Sheehan tomorrow on The Grand Round Up to learn from how they’re thinking about the ACCESS opportunity. Given ACCESS's deflationary goal, inviting new entrants into Medicare seems like the best bet here.
PEPTIDES
Good news for procuring your Wolverine Stack
On Wednesday, a federal notice went out indicating that the FDA is considering moving towards easing restrictions on peptides (including BPC-157 and TB-500), seemingly in part to move the production of these peptides out of the grey market, where quality is understandably a concern today.
Investors and entrepreneurs see a huge opportunity here, as evidenced by Hims & Hers’ stock performance this week. It was up ~50% for the week after the FDA news came out on Wednesday:
Source: Google Finance
Hims is not the only company preparing for the opportunity here, of course. On Friday, Superpower’s CEO took to LinkedIn to announce its new forthcoming peptide offering. A new startup, Protocole, emerged from stealth this week with $6 million in seed funding, designed to offer a more clinically sound approach to peptides.
I feel like we’ve been talking about peptides every week already in 2026, and it doesn’t seem that momentum is slowing down any time soon.
CHART OF THE WEEK
I’m a sucker for a good 2×2 matrix, and found the chart below on the upcoming pharma revolution to be fascinating. It suggests that the next step in pharma is going to be moving beyond molecule development and winning in the employer market as an ecosystem play like LillyDirect is pursuing.
Source: Hardcore Zen Substack
The entire article makes for a fascinating read as it ponders the ideal skillset of the next CEO of PhRMA, the trade association for the pharma industry, and notes that this person has the opportunity to craft a new narrative about pharma, leveraging GLP-1s as the drug that changes the framing, as everyone questions the value of PBMs.
A NUMBER TO PONDER
$163 billion
I find it fascinating to read the article about PhRMA above and this one back-to-back this week. It’s easy to see why a pharma advocate would suggest that the winning play for pharma is as the ecosystem supporting public health. It’s equally as easy to see the challenging tradeoffs at hand when reading the Keytruda article. I also can’t help but think to myself, for as much as people critique the US healthcare system, if I had a loved one who needed Keytruda, I’m pretty sure I’d rather be here than anywhere else.
Other Top Headlines
Abridge announced a content partnership with JAMA and NEJM, and will use the content in its Clinical Decision Support product offering. Strikes me as a very logical move for Abridge to add to its product offering as it and others vie to be the trusted AI support tool for providers. The conversations between AI platforms and large provider organizations must be fascinating at the moment. Take Sutter Health, for example: it announced a partnership with Abridge in March 2024 and is a featured logo on Abridge’s website. In February 2026, Sutter also announced a partnership with OpenEvidence. If you’re Sutter, how do you evaluate these relationships as the product offerings increasingly converge? Does it make sense to have multiple relationships here? It seems worth noting that Abridge’s announcement featured a quote from Sutter’s CEO, while OpenEvidence’s announcement did not. We’ll be chatting with Abridge’s CEO, Shiv Rao, live on The Grand Round Up tomorrow to talk about these content partnerships and what they imply for Abridge’s strategy moving forward.
Baylor Scott & White Health Plan bows out of the ACA and Medicaid markets. It currently serves roughly 2.6% and 3.5% of those populations in the state, respectively, and will cut 321 jobs as part of the move.
Oregon placed ATRIO Health Plans under receivership, as it faced mounting losses and has a mounting unpaid claims backlog. ATRIO currently has ~36k members in Medicare Advantage plans. The state noted that ATRIO lost $50+ million in 2025 and ended the year with only ~$10 million in capital and surplus.
The WSJ highlighted Wakely’s latest ACA enrollment data (see below for more), noting that 14% of the market did not pay their first premium in January, with some states seeing more than 25% of individuals not paying their first premium. Wakely estimates the overall market contraction will be 17% to 26% from 2025, with increased market morbidity that will create uncertainty for 2027 rate filings.
Trilliant Health released a report on the behavioral health market, highlighting data around supply and demand dynamics. The headline is that behavioral health utilization increased 62% between 2018 and 2024.
President Trump signed an executive order this weekend on psychedelics, removing barriers to their use in the treatment of serious mental illness.
Funding Announcements
Keebler Health, an AI platform using clinical data for risk adjustment, raised $16 million. Flare Capital Partners led the round. It has broader aspirations to support a variety of population health use cases.
Joyful Health, a revenue cycle tool that supports provider revenue capture, raised $22 million. CRV led the round. Joyful’s post notes that providers are losing out on $125 billion of revenue they’ve earned each year. Kinda wild that is half a percent of GDP right there.
Ultralight, an EHR for longevity clinics, raised $9.3 million. The General Partnership led the round. Ultralight, which was formerly known as Vibrant Practice, reports being live in 75 longevity clinics today.
Wavelet Medical, a fetal EEG monitoring system, raised $7 million. It is partnered with venture studio Aegis Ventures.
Protocole, a “clinical grade” model for peptides, raised $6 million. Rare Capital led the round. It appears that Protocole is attempting to meet the rising interest in peptides with a more credible offering than the grey-market brands currently on the market. It seems like a logical bet if you think peptide consumption is going more mainstream.
What I’m Reading
A Wakely report dives into the latest ACA enrollment data, highlighting the shift in enrollment across metal levels (Bronze & Gold = up, silver = down). It expects average morbidity to increase by 2.9% to 6.5% in 2026 across the market as a whole, while noting that is an average, and specific markets will see swings outside of that range, suggesting that we will need to keep a close eye on 2027 rate filings.
An article published in Nature Health analyzed 500,000 de-identified health-related queries of Microsoft Copilot to better understand what people are asking chatbots about. It highlights some interesting different usage patterns across desktop and mobile searches, which seem aligned to people using desktop at work and mobile for personal searches.
The Purchaser Business Group on Health’s annual survey of ~two dozen large employers is out, highlighting affordability, transparency / data analytics, and advanced primary care as the top three priorities. The data also notes that 27% of its members are already using a non-traditional PBM, which strikes me as surprisingly high. That would seem to imply that 11 of the 40 largest employers in the US have moved on from a traditional PBM? If that’s true, I’m a bit confused about why so much energy is being spent on breaking up the traditional PBMs when the market already appears capable of handling it on its own.
A Sherwood News report highlighted that the DOJ submitted a new motion this week in a case against GLP-1 telehealth startup Zealthy, which now appears to be nearing bankruptcy. Zealthy was founded by the co-founder and CEO of Cerebral, Kyle Robertson, and was originally accused of deceptive conduct alongside Cerebral and a few other related entities in 2024. Robertson has a pretty remarkable track record at this point, and not in a good way. Reading stories like this, I wish the legal system worked more swiftly at times.
Bob Wachter highlights on Substack how Epic threatens the AI market, leveraging its market dominance to win at health systems over better startup alternatives, and how that might thwart innovation in the market. While I certainly understand the conceptual concern, I look at the rapid growth of companies like Abridge and OpenEvidence and scratch my head a bit. If Epic’s market dominance is supposed to be thwarting AI innovation, they’re seemingly not doing a very good job of it at the moment.
This Week on HTN
Community Brain Trust: MA final rate notice, the unbundling of maternity care payment codes, unpacking MEDVi and the broader GLP-1 landscape, and more. Read here.
Weekly Health Policy Briefing: Why I like ACCESS even though I'm skeptical it will work and state level insurance experiments. Read here.
The Grand Roundup: CMMI’s LEAD program and engaging specialists via CARA, MA final rates and benefit cuts, Teladoc’s valuation conundrum, AI creating confusion in private markets, and SimpliFed’s $10.8M Series A to extend OB care.
The case for investing in maternity care and the driving forces behind SimpliFed’s $10.8M Series A | Andrea Ippolito (SimpliFed)
CMMI LEAD and three key changes from ACO REACH: incorporating specialists, using AI-inferred risk, and simplifying tracks | Gabe Drapos (Pearl Health)
How CMS Administered Risk Arrangements (CARA) bridge the gap between ACOs and specialists | Will Gordon (Manatt Health)
An investor’s view of the private market, and navigating AI-driven uncertainty | Conor Green (Truehelm)
Inside alternative plan design: the mechanics and behavior change driving employer cost savings | Craig Allen & Nancy Wang (Sidecar Health)
Featured Jobs
Managing Director of New Ventures at Redesign Health, a venture and applied technology firm. Learn more.
Hybrid (SF or NY)
Managing Director of New Ventures - Cedars-Sinai at Redesign Health, a role dedicated to the new joint Digital Innovation Platform with Cedars-Sinai. Learn more.
Hybrid (LA)
Chief of Staff at Fortuna Health, a Medicaid navigation platform. Learn more.
$160K - $190K | On-site (NYC)
Operations Analyst at Two Chairs, a virtual mental health provider.. Learn more.
$98K - $133K | Remote
Director, Value Based Care Operations at Duet, a national network of Nurse Practitioner-owned primary care clinics. Learn more.
$150K - $170K | Hybrid (NYC)
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.