I Built My Own Personalized Panel of Renowned Docs for $20/month Complete With Customized Protocols

A few weeks ago on our weekly Food is Health Livestream, we sat down with Pryce Ancona, an industry leader in healthcare interoperability (and friend!) who has spent the last decade connecting the disconnected systems of American healthcare (or at least trying). What started as a conversation about frozen olive trees in Texas turned into an hour-and-a-half deep dive into how AI is about to fundamentally change the relationship between you and your health data. Whether healthcare believes it or not. And guess what… our guest walked away rethinking the true implications.Here’s the full video. If you somehow find this TV worthy, you can now access our content on the Substack TV app on your smart TV. FYI, it took a hot second in our house, one TV let me download the app and one didn’t. Because some folks prefer the written word, we have a recap below.If you are new here - WELCOME! And thank you, there are thousands of you as of February. Just a little context for those who have just jumped in…when we talk about Food Is Health, we’re talking about a full systemic redesign where nutrient density and nutrition becomes affordable. From soil, through ingredients, through packaged goods, all the way through distribution, retail, restaurants, and into healthcare itself. To break it down, we created a framework we call System A (inherent health - evolution - pre-20th century - not scalable), B (current state - cheap calories fueling a chronic disease epidemic leading to a crippling healthcare cost burden). And System C - our mission - a new system built on first principles, designed around human outcomes and human health. Our current healthcare system still handles what it was meant to do (acute care), but a parallel system emerges that’s focused on keeping people healthy in the first place starting at the root through food. Fully upstream with primary prevention.Meet Pryce: The Guy Who Connects Healthcare DataI first met Pryce Ancona when I pretty much forced him to share an UBER to the airport for when we were both leaving the first HHS/CMS Digital Tech Bootcamp in DC and became an immediate fan of System C. He’s now that friend who texts me randomly when he sees something that fits into the future state we aim to build and brings me a massive smile. Pryce started his career at 21, fresh out of college, working for a massive electronic health records company in Wisconsin - one that everyone in health tech has heard of - you know the one. Over the past decade, he’s become one of the sharpest minds in healthcare interoperability - which is a fancy way of saying he figures out how the hundreds of thousands of disconnected healthcare data silos in this country can actually talk to each other. Or at least tries REALLY hard.He’s now at HTD Health, lured there by our friend, the man, the myth, the legend, Brendan Keeler - who writes a brilliant Substack (Health API Guy). The firm consults with payers, providers, and software vendors on how to connect systems, interpret info-blocking regulations, and build strategies for data mobility. When he describes his work, Pryce puts it simply: we made all these software systems, and now we’re trying to make them talk to each other.And that brings us to the big news.What Actually Changed: AI Meets Your Health RecordsHere’s the shift that Pryce highlighted: OpenAI and Anthropic have both announced that their AI systems can now talk directly to your doctor’s systems. OpenAI is using b.well, and Anthropic is using HealthEx.Think of it like Plaid for healthcare. The same way your budgeting app uses Plaid to connect to your bank accounts, companies like HealthEx and b.well allow apps - and now AI tools - to ping your health system’s API and say, “I have one of your patients here. I can prove it’s them. Can I have their chart?” And thanks to government mandates around info-blocking, certified electronic health records have to respond.This is the foundational infrastructure shift. Your health data is becoming mobile.Now, in practice, we’re still in the early innings. When I connected through HealthEx, it only pulled a tiny sliver of my actual health data AND that was separate from the Connector via Claude which found nothing. But according to HealthEx that will be changing quickly - and after chatting with their bada$$ CEO I believe her/them. In my case, during these early innings, I still had to go to Quest, LabCorp, and almost every other provider to manually download PDFs. It took me three to four hours to pull together ten years of data across 40 documents. I was a dog with a bone once I got started.But here’s the thing: that friction is temporary. The data pipes are being built right now.The “Who Owns the Data” QuestionCarter pushed Pryce on a key architectural question: Are we moving toward a world where you own your data and share it back to the system, rather than the system owning it and giving you access? Case in point - my own experience. It’s why I was willing to dedicate the 3-4 hrs because I knew it would always be there.Pryce’s answer was unequivocal: Yes. That’s literally what we’re working on.The government is going all-in on patient empowerment. CMS launched the Health Tech Ecosystem initiative, asking networks, EHRs, payers, and patient-facing apps to pledge toward data mobility. The whole policy direction is making it so that if a provider blocks the exchange of your healthcare information, there are financial penalties.And just last Friday CMMI (Centers for Medicare and Medicaid Innovation) leadership was quoted as saying": Develop NOT for RPM economics, but leverage LLMs to help patients self-manage. If you've already got something built, think about the automation and self-management side. - CMMI LeadershipBut Pryce also named the tension honestly. There’s still not a clear economic incentive for many providers to actively help patients take their data elsewhere. An object not in motion tends to stay not in motion, as he put it. The incentives are changing, but slowly. Again I’m brought back to the new CMMI/CMS ACCESS program (first round applications due in April for July 1st Go-live). ACCESS is an Outcomes Based Payment, Direct to Consumer, Original Medicare program intended to aid eCKM (early cardio kidney metabolic) and CKM (cardio kidney metabolic) along with MSK (musculoskeletal) and BH (behavioral health). Once the rates were released last Thursday it became clear via tech, not humans. BUT interestingly enough, the figures are also enticing to CPGs and grocery and there are some great white label options to meet the tech/Part B requirements. I’m not ready to let the fat lady of tech bros sing on this one yet. We can do better than those inevitable dumpster fires. Carter’s reframe was sharp: if Epic is focused on acute care, and standard care is deficient 80% of the time, then there’s a massive population of people who could take proactive steps to never develop chronic disease. Who has pole position in that story? Nobody has an incumbent advantage there. It’s wide open.But I always say, the food industry could take a TON of marketshare from healthcare if they so choose.The Oura Ecosystem: An Alternate Healthcare UniverseBefore we dove into the Claude reports, I wanted to make a point about OURA that I think the whole healthcare industry needs to hear.I’ve gone deep as a power user myself of the OURA ecosystem, and what I found is that they have quietly built an alternate healthcare delivery system. This isn’t just a fitness tracker for the worried well. Look at what’s inside the app: you can order a continuous glucose monitor (Stelo - delivered in some places same day via Amazon), track your blood sugar data alongside your sleep and heart rate variability, order health panels through Quest (schedule the appointment, pay for the labs, get results - all within the Oura app), and receive personalized health recommendations based on your data.They’ve built a direct-to-consumer healthcare ecosystem that controls the patient relationship the same way a health system does. And if we look at the ACCESS program once again, come July if they become an ACCESS participant, have patients attributed to them.Unfortunately, the general consensus is that their health nudging is generalized LLM at best. A far cry from what I was able to generate with a $20/mth subscription to Claude Pro which is a bit humorous given the monthly OURA subscription is almost half of the cost of Claude Pro ($20/month).Is OURA perfect? No. I think their behavioral nudging needs serious scientific rigor - they should partner with someone like our friends Marten den Haring & Dr. Amy Bucher over at Lirio. But the infrastructure is there. And here’s my prediction: I think they’re going to put rings on every senior’s finger come July 1st and pull as many people as possible into this ecosystem. When I saw the rates on Thursday night, it felt like Tom Hale wrote up what he needed and handed to Oz over dinner one night.And the behavioral piece? It works. I shared my own OURA data on the stream - how I’ve been struggling with sleep and nervous system recovery, how installing a home sauna finally allowed me to nail my evening routine, and how one night of doing it right brought my resting heart rate back down to normal, delivered my best sleep score in weeks, and produced an all-time high heart rate variability for the year.Although our Unconference last week and my time leading up to it at altitude climibing mountains seem to have sent my resting heart rate into a uproar.That’s the behavioral nudge loop in action. You measure, you learn, you change, and the ring tells you it worked (or it didn’t). Even our favorite doc - Graham Walker, MD shared with me yesterday that his early experience with his OURA so far is favorable. And yes, OURA is completely inaccessible for the majority of the population. But if my prediction is accurate, not for long. I mean, they’ve already operationalized deals with Medicare Advantage plans like Essence.I go deep because I learn by doing and I’m determined the components of System C exist they just need to be identified and integrated with their long lost missing pieces.Claude for Healthcare: The Reports That Blew My MindNow the main event. I loaded 40 .pdfs & .pngs into a Claude Project - 10 years of health records from Quest, LabCorp, my OB-GYN, a DEXA scan, brain imaging data, EEG reports - i also loaded biopsy results but those didn’t seem to be picked up. My project instructions were one sentence: “I am using this project to understand how to optimize my health by leveraging my actual health data that is shared with Claude and my desire to focus on healthspan using the definition Peter Attia uses.” That’s it. (Note - despite the controversy surrounding Attia currently, I still think the healthspan framework is brilliant)Report 1: The Four Horsemen Assessment & Health Optimization ReportMy first prompt was simple: “using what you know about my health data can you give me an overview of what I should do to be the healthiest version of myself”. Claude took all of my data, cross-referenced Peter Attia’s framework, and produced a comprehensive report organized around the Four Horsemen (cardiovascular disease, metabolic dysfunction, neurodegenerative disease, and cancer).I didn’t ask it to use that structure. I didn’t ask it to rate me. It figured it out on its own. It gave me an executive summary, told me flat out that I have a thyroid problem that needs urgent attention, connected the dots between my thyroid and cognitive test results, and suggested that before I consider something like a hyperbaric chamber for my brain concerns, I should get my thyroid sorted out first.But the detail it provided beyond that, the biometric by biometric break down was brilliant. It didn’t use the “proprietary” algorithm of biological age to hold me hostage to a monthly subscription, instead it gave me the brass tacks break-down. It gave me detailed intel into my health and even produced a protocol that tied directly to the four horsemen and healthspan. It began to finally give me a roadmap of (genetics aside) how to be able to be climbing mountains and lifting suitcases into the overhead bins at 80 instead of waiting for wheelchair assitance. But it did create a level of concern about my brain health. And instead of doing what many are doing right now and running to my doctor to get clarity (Graham had his own experience with this with DexaFit recently), I decided to have it dig deeper.Report 2: The Deep Dive on Brain HealthBecause the first report flagged neurodegenerative concerns (my mom is dying prematurely from dementia so it sent shockwaves of fear), I went deeper. One prompt: “an you go back through all of the files i uploaded and take into consideration all of the new labs, the beam report about brain compared to the MRI from prenuvo, add the dexa and also consider the vo2 max report and update thing”Claude pulled together my Pranuvo full-body MRI data, EEG results from a 2018 scare, and cognitive test scores, and produced an executive summary that no neurologist has ever delivered to me - connecting verbal memory patterns, reaction time deficits, and thyroid dysfunction in a way that actually made sense.The level of insight was mind blowing. Of course I sent it over to our friend Dr. Dan Durand over at Prenuvo and we are scheduled to dig in with his team in a few weeks. Don’t worry we will try to get them to come onto a Livestream.Report 3: The Brain Optimization ProtocolThen I asked it to take that scary brain report and tell me how to fix it, incorporating the work of some of my favorite Neurologists. One prompt: "what are suggestions especially from the Sherzai’s, Dr. Dale Bredesen and Valter longo about how to specifically improve brain size and health taking into consideration my specific weaknesses above". I warned you it was basic!Claude produced an integrated protocol that showed what each doctor would prioritize, flagged that fasting-mimicking diets are tricky with a thyroid problem (I already knew this, but it independently caught it), and ended with a unified action plan. It was like assembling a panel of three world-renowned clinicians and having them review my case together.Three prompts. Three reports. The kind of analysis that would cost tens of thousands of dollars in specialist consultations - if you could even get them to talk to each other.“This Is the Worst It’s Ever Going to Be”Back to the phrase that stuck with all of us: This is the worst it’s going to be.The data accessibility will only get better. The AI will only get better. And the people using these tools - especially Gen Z, who are essentially LLM-first in how they understand the world - will only get more sophisticated with their prompts.If a perimenopausal woman like myself who can’t find a doctor to help can produce these reports with three one-line prompts using today’s tools and today’s clunky data collection process, imagine what this looks like in a year or 2 or 5. Imagine what it looks like when your complete health record flows seamlessly into an AI that knows every published protocol and can personalize it to you - which if HealthEx keeps after it and TEFCA helps us out, should be much sooner than we thought possible.Pryce, who started the conversation as a self-described AI skeptic who “still rolls his eyes at LinkedIn posts,” left the stream saying this was one of the first AI applications where he thought: Okay, this is not just a better search engine. This is something fundamentally different.His analogy was beautiful: as an avid gardener, he’s always noticing how similar humans are to plants. Your circulatory system is your root system. If you keep feeding it crappy soil, you might not see it in the plant right away, but it’ll be less prepared for the next cold snap. What I’d done with Claude, he said, was:Make the root system visible - and then bring in the world’s best gardeners to help me tend it.The Big Question: Where Does the Disruption Land?We don’t know yet exactly where the disruption will show up. Right now Adam Carewe MD would tell you it’s landing in his virtual care waiting room, currently full of patients who just got their latest Function Health results and are in a tizzy. Carter Williams raised the grocery store example - pharmacy teams within grocery chains who know food matters but still think in pharmaceutical terms, who want to reimburse patients through pharma rather than reimagining the food delivery system itself. Nobody’s even talking to the CPGs about reducing diabetes risk through better products.It’s all swirling. And in that swirl, there’s a wormhole for the good guys - the people who want to make this about human outcomes, not continuing to monetize disease. That’s what we are building in System C.Maybe OURA’s data will expose a market of 50 million people who care about their health, and grocery stores will realize they should redesign their shelves so that it’s hard to make a bad decision. Maybe the system reshapes itself so that all the bread is good, all the brownies are good for you, and all the vegetables have better nutrition.As Carter put it: everybody’s above average. That’s the Lake Wobegon effect at the grocery store. And that’s what System C could look like.As Pryce said at the end: “I’m tired of payers being the purser and everything being preventative or acute instead of just actually taking care of the body.” He came in skeptical (well as skeptical as the ever chipper Pryce can be…). He left asking whether he should get a glucose monitor and an MRI.That’s the shift. And it’s just getting started.Cliff NotesFor those who prefer the highlights:* AI + Health Records = Now: OpenAI (via b.well) and Anthropic (via HealthEx) can now connect directly to your health records through interoperability infrastructure. The plumbing is real, even if it’s still early.* Data Ownership Is Shifting: Government policy is pushing toward patient-controlled data. Info-blocking penalties are real. The momentum is toward you owning your health data and sharing it back to the system.* Oura Built a Healthcare Ecosystem: Beyond sleep tracking, Oura has glucose monitoring, lab ordering through Quest, health panels, and behavioral nudging — all in one app. They’ve quietly built a direct-to-consumer alternative to traditional healthcare delivery.* Claude + Your Data = Your Own Medical Panel: With 40 PDFs and three simple prompts, I generated integrated health reports that rival (and in some ways surpass) what specialist consultations costing tens of thousands of dollars have produced.* “This Is the Worst It’s Going to Be”: Data access will improve, AI will improve, and user sophistication will improve. What’s possible today with manual effort and basic prompts will be seamless and ubiquitous within five years.* System C Needs People at the Table: The Unconference in Denver (Feb 11-13) gathered the right minds together - food, health tech, policy, AI — to design the system that puts human outcomes first. Come with your health data downloaded. Get full access to Food is Health at foodishealth.substack.com/subscribe