Kristen Valdes, Founder & CEO of b.well, and Janneta Tabakov, Enterprise Growth Leader at Perplexity, sat down for the latest Discuss & Disrupt webinar to talk about where healthcare is headed and why the home is at the center of it. Here’s what you need to know.
The Connected Home Has a Blind Spot
Connected home technology has become the norm. The average U.S. household now contains more than 20 connected devices, and that number is only growing. Ring cameras, smart thermostats, sleep trackers, air purifiers, wearables, smart refrigerators. All of them generate data about how we live, and yet none of them know anything about our health.
That’s the gap Kristen Valdes, Founder & CEO of b.well, and Janneta Tabakov, Enterprise Growth Leader at Perplexity, explored in our latest Discuss & Disrupt webinar. As Kristen put it, “Your home today knows your thermostat preference. It has no idea that you’re on blood thinners.”
It’s a simple statement that captures a massive problem. The home is already rich with behavioral data. It knows when you wake up, when you go to sleep, and how active you’ve been, but without health context, it’s optimizing for comfort rather than care. The $180 billion global smart home market is generating enormous amounts of data, while an estimated tens of billions in preventable harm occurs annually from things like medication errors and poor care connectivity, problems the home could actually help solve if it only knew.
The Missing Layer Is Finally Here
The reason health data has been absent from the connected home is structural, not technical. Health records have historically been locked inside fragmented, siloed systems that don’t talk to each other, but that’s changing.
Because of FHIR interoperability rules and the information blocking regulations passed in the U.S., individuals now have the right to collect their complete longitudinal health record through any third-party app of their choosing, including frontier AI models. b.well has demonstrated the ability to collect more than 96% of an individual’s medical record over the most recent five to ten years since health records became digitized in the U.S.
Why does completeness matter so much? A recently published peer-reviewed study from Google Research answered that directly. When both clinicians and AI evaluators rated health responses, AI with access to a patient’s complete personal health record was rated more than twice as helpful as AI without it. Independent clinicians preferred AI-generated responses when the AI had full health context. The more complete the data, the smarter the AI, and in healthcare, that gap between smart and not smart can mean real harm.
From Chatbots to Agents
Understanding why 2026 is the moment this changes requires understanding a shift in how AI tools work. For the past several years, AI has operated primarily in a chat paradigm — you ask a question, you get an answer. That model is giving way to something more powerful, agents.
Janneta Tabakov demonstrated this live using Perplexity Computer, a product that orchestrates a team of autonomous AI agents working in parallel. Rather than answering a question, Perplexity Computer takes a goal and executes it from start to finish. Ask it to build a weekly health dashboard for your family, and it pulls in data from wearables, medical records, and connected home devices and delivers exactly that. The result isn’t a response. It’s a deliverable.
In the demo, Perplexity Computer connected to synthetic EHR data through b.well, alongside tools like Apple Fitness and Oura, to generate a whole-family health snapshot for the week. It flagged a missed dentist appointment, surfaced a potential medication-food interaction before breakfast, and provided focus areas and a suggested grocery list, all without the user having to ask for any of it. As Janneta described it: “Think of English as the new programming language.” You give it a task in plain language, and it builds.
This shift from asking to delegating is what makes the connected home finally viable as a health tool. The home doesn’t need to wait for you to notice something is off. It already knows.
Real Use Cases, Not Hypotheticals
One of the most grounding moments of the conversation was Kristen’s walkthrough of Warfarin — one of the most commonly prescribed blood thinners in the world, taken by millions of patients for conditions like AFib and clotting disorders. Warfarin’s effectiveness is directly tied to dietary vitamin K intake. Eat more leafy greens, and the drug loses effectiveness. Eat less, and bleeding risk spikes. It’s a delicate, ongoing balancing act that cardiologists manage, but patients navigate daily, largely on their own, between appointments.
The connected home already sees what’s in your refrigerator, what’s on your shopping list, and what your dietary patterns look like. When that data is combined with a health record indicating this person is on Warfarin, it becomes genuinely protective — not a one-time flag, but a continuous read on whether your diet is staying consistent with your prescriptions.
That’s one example. The same principle applies across the home: a sleep tracker that connects a dip in sleep quality to a medication your doctor recently adjusted, an air purifier that knows your allergy labs came back elevated and automatically increases filtration, an overnight monitor that surfaces a pattern change in your oxygen saturation in a pre-appointment summary you can take directly to your doctor. These aren’t futuristic scenarios. The technology exists today. What was missing was the health data layer, and that’s no longer the case.
Trust Is the Foundation
A connected home with access to your prescriptions, labs, wearable data, and family health history is powerful. It’s also a significant ask of consumers. Both speakers were direct about what it takes to earn that trust.
For Perplexity, the commitment is explicit. Health data is never used to train AI models, never sold to third parties, encrypted at rest and in transit, and always revocable by the user. As Janneta explained, “trust is the product, because if we don’t get trust right, nothing else really matters.”
Kristen pushed further on a common misconception in the market that HIPAA equals privacy and security. It doesn’t. Third-party apps governed by the FTC actually face substantially greater exposure and oversight and as she noted: “Everyone monetizes your health data but you. The challenge is that under HIPAA, you don’t actually know when your health data is being shared.” Apps operating outside HIPAA, like b.well and Perplexity, are required to be explicit about real terms of service, provide clear consent flows, and offer a genuine right to be forgotten.
Trust isn’t only about privacy; it’s equally about accuracy. That’s the reason b.well spent a decade building its 13-step data refinery, normalizing data across nearly a thousand different EHR systems, resolving duplicate medications, standardizing units of measurement, and applying the right coding to make health information readable by AI in plain language. Incomplete or incorrect health information, as Kristen put it, is more dangerous than no data at all. Verified data is the prerequisite for trustworthy AI.
The Connected Home Is Already Here
The infrastructure is in place. The interoperability rules exist. The agent technology is working. What Kristen and Janneta demonstrated in this webinar isn’t a roadmap; it’s a preview of what’s already deployable.
For home device makers, insurers, and health systems, the implication is that the businesses that win in this next era won’t be the ones with the most sophisticated models. They’ll be the ones with the best data and the most trusted relationship with the user. “Your AI is only as good as what it knows about the person it’s serving,” as Janneta put it.
The connected home is about to become the most important place in healthcare. Not because of any single device, but because of what happens when all of them, finally, know who you are.
