> ## Content Index > Fetch the complete content index at: https://marginsofcare.com/llms.txt > Use this file to discover other available public pages before exploring further. # Oura’s Data Moat, Agentic Interfaces, Tandem Health Goes Agentic - URL: https://marginsofcare.com/ouras-data-moat-agentic-interfaces-tandem-health-goes-agentic/ - Published: 2026-09-14T19:55:42.000Z - Updated: 2026-09-16T19:08:32.000Z - Author: Eva Worst Scott Galloway from Prof G Media [wrote about](https://www.profgmedia.com/p/one-ring-to-rule-them-all) the Oura ring (emphasis mine): > Similar to the Apple Watch, the Oura Ring tracks 50+ health and fitness metrics. Unlike the watch, which a wearer often removes at night for charging, the ring goes more than a week on a single charge and is worn nearly continuously. According to the company’s S-1 filing, the median subscriber wears their ring 23 hours per day. **By manufacturing the device, continuously collecting data, and adding a thick layer of technology and innovation, Oura may be collecting more real-time, first-party data on wellness than any business in the world.** CEO Tom Hale positions the company’s product not as a wearable, but as a “health intelligence platform.” He continues: > The Big Four are in the health and wellness space, to varying degrees, but their SoHI (significant share of the health interface) is limited by their business models. Engagement time is a decent proxy for SoHI. Amazon and Meta track users on their respective platforms and, with permission, across the web, but their insights are derived from the content we consume and the things we purchase. Apple uses privacy as a wedge to decouple its brand from the data-harvesting practices of its competitors — a strong PR strategy that throttles what the company knows about its customers. Apple supplements its SoHI with iPhones and watches, but neither device has a 23-hour-per-day lock on the user. And AirPods outsell Apple Watches by more than 2:1, making the watch an also-ran wearable inside the company’s ecosystem. **Meanwhile, a ring has greater proximity to blood flow than a wristwatch, making it the more accurate tool for measuring vitals.** Google comes closest to Oura’s SoHI, as it tracks users online, on Android phones, and via their Fitbits (Google acquired the fitness tracker in 2021). Not only is Oura simultaneously tracking more vitals that any other company in the world, it is also doing it 23/7 and with the best possible non-invasive proximity. Lucas Mittelmeier made another [great connection](https://healthtechofftherecord.substack.com/p/ouras-95-problem-digging-into-their) linking technology accuracy, standard of care maturity and business model focus (emphasis mine): > **Many of the newer tracking devices (rings, wristbands, micro-needle patches) operate on inherently inaccurate technology.** Most of them are in the business of estimating, especially the non-invasive ones, and are likely tempted to oversell their accuracy. Pragmatically as an investor, I’d check a) how strong is the standard of care and b) whether they focus on consumers or a medical market. **The weaker the standard of care and the more it’s a consumer play, the better your odds of success.** Oura is the living proof. He also did an overview of Oura's [startup to IPO journey](https://healthtechofftherecord.substack.com/p/who-will-make-the-most-in-ouras-ipo), if you want to dig deeper on the company. ### Agentic Interfaces Today, when a patient gets sick and goes to the doctor, the doctor performs an initial triage and begins the [clinical workflow](https://marginsofcare.com/the-ways-of-the-workflow/), which requires multiple touch points with multiple systems, interfaces and people, then analyses all of the input and assigns a treatment. [It’s chaotic, slow and inefficient.](https://marginsofcare.com/cost-of-care-drivers/) We first started applying AI to the individual steps within the workflow but are [increasingly getting to a stage](https://marginsofcare.com/the-ways-of-the-workflow/), where the majority of the interaction is aggregated into a single multi-functional interface: the agentic clinical interface. [90% of clinicians](https://marginsofcare.com/prof-valmed-wins-dmea-award-certify-it-and-they-will-come/) use LLMs in a single-player mode to synthesise the data that comes to them, and the ChatGPT's read-only access to Epic is a [clear sign](https://marginsofcare.com/chatgpt-connects-to-ehr-epics-ai-game/) this will eventually happen to the EHR / ePA layer, which will not just unlock more data, but also unlock the collaborative multi-player mode between different clinicians. Here is a hand-drawn visual representation of the Agentic Interfaces: ![Agentic Clinical Interface / Agentic Patient Interface](https://storage.ghost.io/c/31/96/3196e6e7-a692-4a81-823d-4b6678f19732/content/images/2026/09/Agentic-Interfaces.jpg) A similar thing is happening on the patient level. Patients are tracking various wellness metrics via different apps and wearables. But there is again an increasing aggregation of those data streams into a single interface: the agentic patient interface. LLMs are again very well positioned to capture the market, but wearable companies like Oura and Apple have equally good chances and arguably more trust to carry them. ### Tandem Health’s Opportunity Tandem Health’s 100m EUR investment fits neatly into that reality. CEO Lukas Saari [wrote on LinkedIn](https://www.linkedin.com/feed/update/urn:li:activity:7505139529599860736/): > We are now doubling down across European markets to accelerate the adoption of AI. > Building out the AI-native Clinic Operating System. With an AI Medical Assistant supporting clinicians before, during and after each patient visit. A Patient Flow solution supporting patients before and after their visits. > Working in tandem with clinicians, patients, systems and nations. Like [many](https://marginsofcare.com/prof-valmed-wins-dmea-award-certify-it-and-they-will-come/) | [others](https://marginsofcare.com/epics-eu-opportunity-doctolibs-threat/), they started with an AI that addresses a specific step from the workflow and are now expanding with the aim to orchestrate the entire workflow. This makes sense even as it poses the question to what degree are comfortable aggregating all of our sensitive data into a single company?