@andrewarruda Did you see @CloverHealth presentation? Excited for the future.
English
Da_Pupul
413 posts
it’s a great time to be building in health tech
CMSGov@CMSGov
Happening now! Join CMS virtually for the Health Technology Ecosystem: Live! First Wave Launch. Innovators from across health care will be showcasing the Minimum Viable Product of their new health tech. Don't miss the opportunity to get a first look at the cutting-edge apps designed to transform healthcare delivery. Join now: go.cms.gov/4c9Kgog
English
Special Interest Media@thoughtson_tech
If Cleveland Clinic is using an LLM to sit between clinicians and their data, the real question isn't whether it works, it's who owns that layer long-term. The article linked below argues that Apple's ~$2B grab of Q.ai shows the interface fight has already moved past voice. Health systems like HCA have proven physicians will pay $200-400 a month just to stop typing, with no new clinical insight added. That's a strong signal that the interface itself is the product, not the model behind it. Language, though, may be a ceiling. If smell, gaze, and pre-vocal signals carry more clinical data than any spoken note, then LLMs are a bridge, not a finish line. Cleveland Clinic's pilot fits the pattern, but the firms that lock in the interface layer before the next shift will be very hard to displace. onhealthcare.tech/p/the-interfac…
QAM
Berci Meskó, MD, PhD@Berci
Generative AI (in the form or large language models) will act as the ultimate interface between healthcare professionals and a range of AI-based and other digital tools. Here is another proof of my prediction. Cleveland Clinic is the first health system to pilot "Chart Chat for Nursing", the first EHR-integrated AI conversational tool for inpatient nurses. "𝑇ℎ𝑒 𝑡𝑜𝑜𝑙 𝑝𝑢𝑙𝑙𝑠 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝐸𝐻𝑅 𝑡𝑜 𝑎𝑙𝑙𝑜𝑤 𝑛𝑢𝑟𝑠𝑒𝑠 𝑡𝑜 𝑎𝑠𝑘 𝑞𝑢𝑒𝑠𝑡𝑖𝑜𝑛𝑠 𝑜𝑛 𝑝𝑎𝑡𝑖𝑒𝑛𝑡𝑠’ 𝑐ℎ𝑎𝑟𝑡𝑠, 𝑤𝑖𝑡ℎ 𝑒𝑎𝑐ℎ 𝑟𝑒𝑠𝑝𝑜𝑛𝑠𝑒 𝑝𝑎𝑖𝑟𝑒𝑑 𝑤𝑖𝑡ℎ 𝑓𝑢𝑙𝑙 𝑐𝑖𝑡𝑎𝑡𝑖𝑜𝑛𝑠, 𝑡ℎ𝑒 𝑐𝑜𝑚𝑝𝑎𝑛𝑦 𝑠𝑎𝑖𝑑 𝑖𝑛 𝑎 𝑝𝑟𝑒𝑠𝑠 𝑟𝑒𝑙𝑒𝑎𝑠𝑒. 𝑁𝑢𝑟𝑠𝑒𝑠 𝑐𝑎𝑛 𝑎𝑠𝑘 𝑝𝑙𝑎𝑖𝑛-𝑙𝑎𝑛𝑔𝑢𝑎𝑔𝑒 𝑞𝑢𝑒𝑠𝑡𝑖𝑜𝑛𝑠, 𝑝𝑢𝑙𝑙𝑖𝑛𝑔 𝑓𝑟𝑜𝑚 𝑝ℎ𝑦𝑠𝑖𝑐𝑖𝑎𝑛 𝑝𝑟𝑜𝑔𝑟𝑒𝑠𝑠 𝑛𝑜𝑡𝑒𝑠, ℎ𝑜𝑠𝑝𝑖𝑡𝑎𝑙 𝑝𝑜𝑙𝑖𝑐𝑦, 𝑑𝑜𝑐𝑢𝑚𝑒𝑛𝑡𝑎𝑡𝑖𝑜𝑛, 𝑜𝑟𝑑𝑒𝑟𝑠 𝑎𝑛𝑑 𝑟𝑒𝑐𝑒𝑛𝑡 𝑙𝑎𝑏𝑠 𝑖𝑛 𝑠𝑒𝑐𝑜𝑛𝑑𝑠." It seems healthcare professionals are about to "chat" with the EHR systems across specialties and disciplines, just like what patients do at home. Source: fiercehealthcare.com/providers/ambi…
QAM
English
@chamath Agreed. The tech industry needs to stop talking to itself.
Less conferences. Less manifestos. Less “the future is here.”
Go build something for a nurse in Ohio who’s never heard of Sam Altman.
English
If tech leaders don’t organize and get America on their side, the situation on the ground - as seen in the three charts below - will get worse before it gets better.
That, in turn, will tank the US economy since AI is responsible for much of our incremental GDP.
Someone needs to step up.
English
@DiaTSLAPLTR @topsecretstocks Congrats, you sold at the bottom. That is not patience.
English
@topsecretstocks Bro, I got into $CLOV at the IPO in 2021 and purchased my first shares around $10 and still hold some shares. I sold half of my position two weeks ago. I think you can call me patient as well 🙂
English
English
The number people aren't talking about enough: Anthropic's healthcare exposure is still tiny relative to this trajectory.
When a foundation model company's revenue is growing this fast, their enterprise bets become structurally different decisions. Menlo Ventures Anthology Fund sitting on Qualified Health's cap table isn't just a financial play—it's Anthropic purchasing optionality on the governance layer before healthcare becomes a material revenue vertical. At $30B annualized run rate, Anthropic can afford to seed the infrastructure companies that make Claude deployable inside risk-averse health systems. That's not charity. That's channel development at the foundation model layer.
Most people frame foundation model growth as competition with application companies. The smarter read is that at this velocity, the hyperscalers need regulated-domain infrastructure partners more than those partners need them.
Health systems won't run raw Claude. They'll run Claude wrapped in audit trails, clinical governance frameworks, and deployment monitoring. Whoever builds that wrapper owns the relationship. Anthropic knows this.
The $30B number matters for healthcare AI specifically because it signals the foundation model wars are effectively over for enterprise procurement purposes—and the next fight is for the middleware.
prinz@deredleritt3r
Anthropic revenue (annualized): - January 2025: $1B - May: $3B - June: $4B - August: $5B - October: $7B - December: $8B to $10B -February 2026: $14B -March 2026: $19B -April 2026: $30B (WTF???)
English
@thoughtson_tech @CloverHealth built that long before yesterday.
“CMS is excluding diagnoses from unlinked chart review records starting CY 2027. – somebody needs to build all of that yesterday”
English
new satisfying CMS lore just dropped and its basically a startup pitch deck disguised as a 139 page rate announcement
the CY 2027 MA Rate Announcement finalized a 2.48% net payment increase, about 13 billion dollars. sounds fine until you realize the advance notice projected 0.09% and the industry comment section reads like group therapy for actuaries who feel gaslit by OACT. the bump came mostly because CMS backed off the proposed risk model recalibration and stuck with the 2024 HCC model for another year. skin substitute spending distortions were a mess and nobody wanted to eat that in a single cycle
but the real story isnt the topline number. its the operational mandates buried in the document that are going to force plans to buy things that dont exist yet
biggest one: CMS is excluding diagnoses from unlinked chart review records starting CY 2027. roughly 85% of the 88.8 million unlinked CRRs submitted in 2023 couldnt be matched to any encounter even when CMS tried matching on beneficiary, provider, and dates of service within three days. thats a staggering data integrity problem and CMS just made it a revenue problem. encounter data remediation, chart review linking platforms, point of care risk capture tools – somebody needs to build all of that yesterday
audio only telehealth diagnoses are out for risk adjustment too. separate PDP and MA-PD normalization factors are locked in, which creates a distinct analytics market for standalone drug plans that has been underserved forever. new depression screening measure in star ratings. new polypharmacy and opioid safety measures. 11 admin process measures removed so the program refocuses on clinical outcomes. GLP-1 utilization chaos and the BALANCE model starting CY 2027 with no risk corridor relief for plans
every mandate CMS finalizes is a purchase order that hasnt been written yet. the plans know they need to comply. question is just who builds the tools they buy to do it
wrote the full breakdown as a long form essay walking through each policy change and mapping it to specific company categories that need to exist
Link to the full analysis onhealthcare.tech/publish/home
English
@danielisdizzy That’s why he bought the airlines during Covid and soon after sold them because he admitted not knowing the business.
English
Prices matter more than the business.
Warren Buffett keeps selling $AAPL as it’s overpriced, while he’s accumulating $UNH after its collapse.
Same investor. Same principle.
“I don’t have any ability to predict what a stock will do next week or next month. I will buy them if they’re cheap. I will buy a whole lot of them if they’re cheap and I think I really understand the business.”
English
@Ashton_1nvests UNH is a fraud company. They are shrinking and only survive by rate increases and gaming the coding system. Cms is changing the rules and JNH is losing their money machine. CLOV has showed their AI system works, spun out a company counterpart health. Clov is growing.
English
$UNH is likely becoming a top 5 holding for me at market open.
English
$UNH Today’s CMS news was more than just the relief of receiving a 2.48% Medicare rate increase for 2027. There was also regulatory relief where the 2024 risk adjustment model will continue to be used for 2027. Over $13 billion in additional Medicare Advantage payments is expected for participating plan providers.
Some after-hours gain were lost because stock futures are down on Iran concerns, which have nothing to do with the company.
$ELV $XLV $CVS $HUM $MOH $OSCR $CI $CNC $SPY
English
@ParkFamily81818 @NickelDimerBets @VivekGaripalli @CloverHealth @counterparthlth @kevinholub @andrewtoy Never heard it before, you seem to know it well.
English
@SPrintzian @NickelDimerBets @VivekGaripalli @CloverHealth @counterparthlth @kevinholub @andrewtoy you know what they say. ignorance is bliss
English
Special Interest Media@thoughtson_tech
Watched a hospital system spend $4 million and eighteen months building a prior auth workflow tool that their vendor could have sold them for a fraction of that. At the time, that build cost was the vendor's best sales argument. "You'll never replicate this internally." They were right, until now. That shift from precious to abundant hits healthcare harder than most sectors, because so much of health tech's valuation was built on exactly that scarcity. The moat wasn't proprietary data or clinical expertise or FDA clearance. It was just that software was expensive and time-consuming to build, so once a vendor was in, the rebuild cost kept them there. That logic is collapsing fast, and the companies most exposed are the ones whose entire value proposition was "we encoded your business rules into software before you could afford to do it yourself." What doesn't commoditize is what software was always wrapping around: longitudinal patient data, payer contracts, regulatory certifications, and the clinical relationships that make any of it actually usable in practice. Software being free changes everything about the build-versus-buy calculus for large health systems and payers, and almost nothing about what actually makes a healthcare company defensible once the build cost is gone. Wrote about the full breakdown across payers, providers, and pharma here: onhealthcare.tech/p/the-free-lun…
QME
CMS 2027 Medicare Advantage final rates are due TODAY.
Here's the unusual options flow building from my scanner:
$HUM (at $183) — 200 calls: 8.95x vol/OI, 165 puts: 5.63x vol/OI, IV over 114%
$UNH (at $281) — 302.50 calls: 4.26x vol/OI, 280 puts: 4.16x vol/OI
$CNC (at $35) — 30 puts: 4.64x vol/OI
Heavy positioning on both sides. Last time rates disappointed in January, $UNH fell 20%.
Traders are pricing in a big move. We should know soon.
English
@ParkFamily81818 @NickelDimerBets @VivekGaripalli @CloverHealth @counterparthlth @kevinholub @andrewtoy Not my problem.
English
@SPrintzian @NickelDimerBets @VivekGaripalli @CloverHealth @counterparthlth @kevinholub @andrewtoy clearly you and I have a very different definition of "winning". keep on "winning" bro lmao
English