Digital Health Platforms: Verily • Andrew Trister
Engineering Precision Care: Verily’s Strategy and Vision
In this episode of unNatural Selection, produced in partnership with Mass General Brigham for the World Medical Innovation Forum, I sit down with Andrew Trister to explore the evolution of Verily, Alphabet’s precision health company. We discuss Verily’s journey from its early days as an innovation lab to its transformation into a platform company, focused on delivering insights directly into clinical workflows. Andrew shares how Verily is shaping its innovation strategy, with a five-year horizon on the biggest opportunities in connecting health data and driving insights toward precision care. At the heart of this vision is Verily’s objective of building a true health decision platform—one that bridges technology and medicine to redefine how care is delivered.
-
(Auto-generated by Spotify. Errors may exist.)
Andrew Trister, MD, PhD, is the Chief Medical and Scientific Officer at Verily, where he leads the company's research, science, and population health initiatives, providing expertise across Verily's portfolio as the company furthers its precision health strategy. Previously, Andrew se1rved as the Deputy Director of Digital Health and Artificial Intelligence at the Bill and Melinda Gates Foundation, where he led the Foundation's development and investment in digital health and artificial intelligence to address global health inequities. Prior to that, Andrew was a founding member of Apple's health team where he led clinical research and machine learning with Apple Partners. Andrew holds an MD, a PhD in bioengineering, an MSc and BSc in computer science, and a BA in Biological Basis of Behavior from the University of Pennsylvania. He completed his clinical residency in radiation oncology at the University of Washington with additional focus areas in medical and bioinformatic3s.
Host (2:03)
Andrew, it's a pleasure to have you here. Thank you for being on unNatural Selection.
Andrew Trister
Thanks, Nic. Great to be here.
Host
Just to kick off and level set, to give people perspective on what drives your work, could you please give us a sense of what need or impact drives what you do?
Andrew Trister
Well, since I started in undergrad focusing on engineering and pre-med classes, I always had this idea that it might be possible to better care for people using technologies. This...
Andrew Trister (2:29)
...was back in the mid-nineties, and the internet was just starting to come out, and I remember very clearly sitting in a class on signal processing, and thinking, why can't we use the same types of mathematical frameworks to better understand what's happening to people physiologically, or what's happening to the healthcare system when we're trying to give care. And that really has been the driving force throughout my whole career, trying to bring the best of what's possible from a technological standpoint to better care for people in every way.
Host
That is a fantastic motivation. And when we think about what you're doing now, especially at Verily, which is driving precision health, what does that mean in your words, and what does it entail for the future of medicine?
Andrew Trister
Yeah, I think precision health is really the next evolution of what a lot of people called personalized medicine for a long time. I think the notion of personalized medicine always conjured up this idea that you would have, you know, a different drug for every single person. I think precision health starts with the individual, and it says: How can we use all the information we can gather about a single person, or about a population of people that are similar to you, to better understand your risk, to better understand your disease, and to be able to better care for you? And this is in contrast to the current paradigm, which is much more of a one-size-fits-all model, where we use population studies to determine what we should do next for every person. And the hope with precision health is that by adding more information, specifically longitudinal and active measurement of people, we can better target treatments and we can better target risk modification.
Host
And when we think about that, you know, this word "data" comes up a lot, but Verily has a very interesting approach to this because you're using a mix of longitudinal data, you have studies, you have the technology angle of this. Can you tell us how you're using data science, AI, and machine learning to make sense of this massive amount of data to achieve that goal?
Andrew Trister
Yeah, that's really the secret sauce, right? It's easy to say, "Let's gather a lot of data about a person." But then to ask, "How are you going to analyze that data?" is the much harder question. And to do this, we need to bring together a few different areas. One is the science, which is understanding, when we gather data, what is it that we're actually measuring? When we put a new sensor on somebody, what is it that that sensor is telling us? And how do we translate that into a clinically actionable insight? And that, in and of itself, is a whole area of science. The second area, which you alluded to, is machine learning and artificial intelligence, which is the way we're actually going to manage the scale of the data and start to see patterns that we, as humans, cannot see. So, an easy example of this is, we might have, let's say, ten different measurements of a single person's heart rhythm, and all ten of them may look normal, but if we combine them with a machine learning model, we can see a hidden pattern that suggests that a person's actually at risk for a certain type of arrhythmia that a cardiologist could never have seen if they had only looked at one signal.
Andrew Trister (5:01)
And so that's where the idea of the AI and the machine learning comes in. It's to find these hidden patterns that suggest a heightened risk or suggest an earlier diagnosis that you would never have been able to find before. And I think the key piece that Verily brings is: we are connecting this to the scientific study, right? So we don't just put out a new piece of technology and say, "Go use it." We are putting it out in the context of a longitudinal study, gathering the scientific evidence to show that this actually works, and that's the difference.
Host
I love that. The idea that there's hidden patterns that are beyond the realm of human ability to find, that's incredible. And when you think about that, you know, Verily has these longitudinal studies, right? So you have the Project Baseline study, which is an amazing study. Can you tell us a little bit about that, and what you're hoping to achieve with that?
Andrew Trister
Yeah. So Project Baseline is a big vision, right? The idea is, can we create the next generation of reference range for human health? So think about when you go to your doctor, and you get a blood test, and they say, "Well, your cholesterol is X, and the normal range is Y." That normal range of Y was derived 40 or 50 years ago from a small cohort of people, mostly healthy, young, white men, to be quite honest. And that's still the reference range we use for most of medical practice. What we're trying to do is, can we take hundreds of thousands of people, measure them actively, longitudinally, and use the best of technology to redefine what healthy looks like, and what a person's individual trajectory looks like, so that when a person goes to the doctor, they don't look at a population-based reference range, but they look at a personalized reference range, and say, "For this person, in this stage of their life, with this set of health conditions, this is what a change in their data means." And that's really the idea of Project Baseline. And we're doing this in different ways. The flagship is the Baseline Health Study, which has been going on for years now, following thousands of people with a very deep and longitudinal set of measurements, all consented to be used for research. We have other components of Project Baseline, such as the Baseline Register, where we're gathering data from millions of people who have electronic health records to be able to study what's happening at the population level. And then we have the Baseline Platform, which is the actual technology that enables all of this.
Host
I love that. The idea that the reference ranges are changing to more personalized or, as you say, precision, that's incredible. And how does Verily's collaboration with health systems, like the Mass General Brigham System, accelerate these goals, especially in Boston, which is such a hub for life sciences and health innovation?
Andrew Trister
Yeah, I think the partnership with Mass General Brigham, which, you know, we've been working with for years now, is a critical piece of this. You can't do any of this without deep clinical and scientific expertise. And what we get from the partnership with MGB is access to world-class clinicians, world-class scientists, and a patient population that is broad and diverse, that allows us to do these rigorous scientific studies. So, you know, an example of this is, we are working with MGB on some of the largest long-term studies on diabetes and heart disease, where we're trying to integrate continuous glucose monitors, we're trying to integrate passive measurements of people, and then layer in a deep set of clinical measurements to actually see: is there a signal in the data that we can find that's clinically actionable? And it's only by having that deep, clinical and scientific expertise that we can actually make sure that what we're building is clinically relevant and scientifically rigorous.
Host (10:14)
I love that. So it's about connecting the science to the clinical practice, to the real world use, and making sure that there's validity behind the tools.
Andrew Trister
Exactly. And I think a lot of times, the industry has suffered from, you know, great pieces of technology that are not actually relevant for the clinic. Or, you know, great insights that have been published in journals that don't translate to real-world impact. And the partnership with MGB is a real-world validation of what we're doing.
Host
I see. I see. And when we think about the future of diagnostics and early disease detection, how do you see the integration of these wearables and sensors transforming that landscape?
Andrew Trister
Yeah, I think that's the greatest promise of all of this, right? The ability to measure a person continuously, actively, and passively is a game changer for diagnostics. Right now, most diagnostics, most disease detection, is event-based. You go to the doctor when you have a symptom, and they run a test, and you get a diagnosis. With continuous monitoring, we can move from event-based diagnostics to process-based diagnostics. And what that means is, we can find signals of disease before a person is symptomatic. And that, in and of itself, has the potential to dramatically change outcomes. Think about, you know, finding cancer earlier, or finding heart disease earlier. It's only by measuring people continuously and looking for those subtle signals that we can find these things earlier. So, for example, on the Project Baseline study, we're looking at things like subtle changes in sleep patterns, or subtle changes in heart rate variability, or subtle changes in activity that may suggest a person is at risk for an infection, or at risk for a decline in their physical or cognitive health. And it's only when you have that baseline of data, that continuous measurement, that you can see a deviation from that baseline that suggests something is happening. So I think the future of diagnostics is going to be driven by these continuous, passive measurements.
Host
And I can only imagine that the challenges associated with that are massive. And so I'm thinking about privacy and security, and the ethical considerations that go into handling this vast amount of incredibly personal, sensitive health data. How does Verily address these challenges?
Andrew Trister
Yeah, that's a great point, Nic. You can't do any of this without trust, right? And I think that's why we're so committed to doing this in a scientifically rigorous and ethically sound way. A few ways we've addressed this. One is through the informed consent process. We are very clear with people about what we are measuring, why we are measuring it, and how we will use that data. And people have to opt-in to that. And they have to be able to opt-out at any point. We're very committed to that. The second piece is the technology itself. The data is de-identified or pseudonymized at the point of collection, and we have very strict internal policies and procedures to make sure that the data is handled securely and responsibly. We don't sell the data. We use the data to advance research, to build new products, and to create better health outcomes. That's the core of our business model.
Andrew Trister (15:02)
And I think the third piece is governance and oversight. We have an external ethics board, we have internal governance bodies that are constantly reviewing the way we use the data, the types of questions we're asking, and the products we're building. And that is an ongoing conversation. You know, this is not a one-time thing where we said, "We're going to do this ethically," and then we stopped. It's a continuous process of review and refinement to make sure we're upholding the highest ethical and privacy standards.
Host
I love that, and I think that's incredibly important, especially with this kind of data. And I think you've hit on a very, very critical point, which is the informed consent and the agency that the patient has over their own data. I think that is an incredible differentiator. I think you've already answered this to a degree, but as a follow-up, how do you foresee the physician-patient relationship evolving with these new tools and the patient being much more informed about what's happening to them, especially in a world where AI and machine learning are augmenting, and in some cases replacing, certain physician tasks?
Andrew Trister
Yeah, I think it's going to be a net positive for the physician-patient relationship. I think that physicians are going to be able to focus on the things that they're uniquely good at, which is the human connection, the empathy, the contextualization of the data. And the AI and machine learning is going to take care of the heavy lifting of data analysis and pattern finding. Right now, a physician is often acting as a data broker. They're gathering information, they're looking at a limited set of data, and they're trying to figure out what's happening. And a lot of that is time-consuming and often inaccurate because the data is sparse. When you have continuous, longitudinal data, and you have the AI doing the initial triage and pattern finding, the physician can walk into the room with a patient and say, "I see that your sleep patterns have changed, and that your heart rate variability is declining. Let's talk about what's happening in your life that might be causing that." That's a much richer conversation than, "Tell me how you've been feeling since I last saw you six months ago."
Andrew Trister (20:00)
So, I think the physician's role is going to become much more of an interpreter of the data and a partner with the patient to figure out what to do next. And the patient's role is going to become much more active and empowered. They're going to come in with their own data, their own insights, and they're going to be a partner in their care, not a passive recipient of care. I think that's the biggest shift, the shared decision-making that's going to come with this data transparency.
Host
I couldn't agree more. And I love the word "partner" here. I think it's spot on. Now, a slightly different question. You've had a very interesting career trajectory, right? From the Gates Foundation to Apple, to now being the Chief Medical and Scientific Officer at Verily. You've seen innovation from the public sector, from the private sector, and now at this intersection of health, tech, and data. What are the key differences and what did you learn at each of those stops that you bring to your current role?
Andrew Trister
That's a great question. And it's true, it's been a fun trajectory. I think at the Gates Foundation, I learned the immense power of scale and impact in global health. The problems are huge, but the opportunity for impact, even with small investments, is massive. And I think that commitment to impact and scale, that I bring to Verily. Everything we do should be measured by: how many lives can we touch, and how much can we improve the health of populations, not just individuals? At Apple, I learned the immense power of design and usability. Technology can be great, but if it's not easy to use, if it's not beautiful, if it's not seamless, it won't be adopted. And in healthcare, where the friction is already high, it has to be incredibly seamless. I bring that commitment to user-centered design and seamless integration to Verily. And then, at Verily, I've learned the importance of scientific rigor and validation. You can have the best technology in the world, but if you don't have the scientific evidence to back it up, it's just a gadget. So, I think the combination of those three things—the commitment to scale and impact, the commitment to design and usability, and the commitment to scientific rigor—is what I try to bring every day to Verily.
Host
I love that. That's a great way to put those three things together. And you know, you're now at a place where you're looking at the future of health, and you're leading that charge. If you had to look out ten years from now, what is the single biggest change that you predict will happen in medical practice because of the work that you're doing?
Andrew Trister
I think the single biggest change will be the shift from reactive to proactive care. Right now, medical practice is mostly reactive. We wait for a person to get sick, and then we treat them. In ten years, I think medical practice will be fundamentally proactive. We will be using the data, the AI, and the continuous measurement to predict a person's risk and to intervene before they get sick.
Andrew Trister (25:01)
We won't just be treating disease; we will be managing health. And I think that's a fundamental change in the way we think about the entire healthcare system. It will affect everything from how we pay for care, to how we train physicians, to how we build hospitals. It will be a total redesign around the idea of proactive, predictive health.
Host
That's an incredible vision. And when you think about that, what are some of the biggest barriers that you see to that vision, and how are you working to overcome them?
Andrew Trister
The barriers are significant, but they're not insurmountable. I think the biggest one is the regulatory and reimbursement environment. Right now, the system is designed to pay for reactive care, not proactive care. How do we create a regulatory framework that encourages and rewards proactive, predictive interventions? That's a huge challenge. The second one is data interoperability and standardization. We're still in a world where data is siloed in different electronic health records, in different sensors, in different systems. We need to create a seamless flow of data so that the AI can actually work its magic. And the third is cultural change. We need to get physicians, patients, and the entire health system comfortable with this new paradigm of data-driven, proactive health. That takes time, it takes education, and it takes trust.
Host
I see. So, the cultural change is huge, and I think that's a great way to put that. I'm going to take a pause on that because I have a slightly different question that relates to all of this. We talked about your experience at Apple. And I think that's fascinating. You were a founding member of the Health team there. What do you see as the role of consumer technology companies in driving this kind of precision health revolution, and where does that intersect with the role of companies like Verily?
Andrew Trister
Yeah, I think the role of consumer technology companies is absolutely critical, but it's different than the role of a company like Verily. Consumer technology companies are fantastic at democratizing access to health measurement. They're putting sensors on people's wrists, they're making it easy to gather data, and they're creating a level of health literacy and engagement that didn't exist before. They are the engine of data collection for millions of people. The challenge, however, is that they often lack the scientific rigor and the clinical validation to turn that data into actionable insights for the healthcare system. That's where a company like Verily comes in. We take that data, we layer in deep scientific and clinical expertise, we run the randomized trials, we do the long-term studies, and we create the clinical and scientific evidence that allows that data to be used by physicians, by payers, and by the regulatory bodies. So, I see it as a symbiotic relationship. Consumer tech companies create the data, and companies like Verily create the evidence and the software to turn that data into better health outcomes.
Host (30:26)
That is a brilliant articulation of that intersection. I think that's a very helpful way to look at it. Now, you also have experience in a field that I think is going to be incredibly important in the future, which is global health and global health equity, given your work at the Gates Foundation. How does this precision health revolution—which is often characterized by very high-tech, expensive tools—address or potentially exacerbate global health inequities, and what is Verily doing to ensure that this revolution is inclusive?
Andrew Trister
That's a question I think about every day, and it's a huge danger, right? The danger is that the precision health revolution only benefits the people who can afford the most advanced technologies, and that would be a tragedy. So, we have to be very intentional about designing for equity. A few ways we're trying to do that. One is through simplicity and low cost. The most impactful technologies in global health are often the simplest and the cheapest. We're trying to take the insights from our advanced research and distill them down into tools that can be deployed at scale in low-resource settings. An example is, can we take a complex AI model for early disease detection, and can we deploy it on a $50 smartphone in a rural clinic? We're actively working on that.
Andrew Trister (35:10)
The second thing is partnerships. We have to work with governments, with NGOs, and with local health systems that are already on the ground and understand the context. The Gates Foundation taught me that you can't parachute in with a solution and expect it to work. You have to be a partner. And the third is data diversity. We have to make sure that the data that trains our AI models is representative of the entire global population, not just a small, privileged cohort. If our AI is only trained on a small, non-diverse population, it will only work for that population. And that's why our collaborations, including the ones that came out of the Gates Foundation, are so important to us—to make sure that we have a global perspective on data.
Host
That is incredibly insightful. I think that's a great answer. So it's about making sure that the technology is accessible and not just for the developed world, but for the entire globe. And I think that's an admirable goal. And when you think about the future of human health and the challenges of an aging population, how do you see precision health helping to address issues like chronic disease management and aging in place?
Andrew Trister
I think this is one of the biggest opportunities for precision health, right? The challenges of the aging population—chronic disease, cognitive decline, physical frailty—these are all process-based diseases, meaning they don't happen overnight. They happen slowly over years, and there are subtle signals of decline that are missed by the current event-based system. With precision health, we can measure those subtle declines continuously. We can find those signals of risk for a fall, or for cognitive decline, years before it becomes a crisis. And that allows for an early intervention that can actually change the trajectory of that person's health.
Andrew Trister (40:05)
So for chronic disease management, instead of saying, "Your blood sugar is too high, here's a pill," we can say, "Your blood sugar has been slowly creeping up for the last six months, and we see that it's correlated with changes in your sleep and changes in your diet. Let's intervene with a combination of lifestyle and potentially a medication change before you hit a crisis." That's the power of continuous data. And for aging in place, it's about providing the right level of support at the right time. It's not about putting a hundred sensors in a person's home. It's about finding the one or two critical measurements that tell us this person is at high risk for a major event, and then intervening with targeted support. I think this is where precision health can be a massive engine for not just better health, but also for cost containment in the healthcare system.
Host
I love that. The idea that it's a massive engine for cost containment as well is fantastic. I think that's a great way to put it. Now, as we wrap up, I want to ask a question that is a little bit more philosophical. You know, given everything that you've done in your career, what gives you the most optimism for the future of human health?
Andrew Trister
That's a beautiful question. What gives me the most optimism is the integration of data and the human spirit. I think for a long time, we've had this false dichotomy between the technology and the humanity of medicine. And I think what the next ten years is going to bring is a recognition that the best of technology is actually going to enable a greater level of humanity in medicine. Because we will have access to information that we can't even imagine now, right? The longitudinal physiology, every measurement known possibly, and would require a superhuman intelligence to even figure out what does all of this mean?
Andrew Trister (46:57)
So getting back to your point on making better precision around risk and then taking action about that, understanding for myself what it would mean and how can I take something to shift my risk will be the larger part of what we can do in medical practice. And I think it will allow for a greater level of connection between people, between, whether that's physicians or nurses, you know, other extenders, or whether it's a person just seeking help from others, how they can connect should be even greater than it is now.
Host (47:34)
I totally agree. Yeah, I think it'll enable richer conversations with a more informed patient that has agency over their own healthcare. And it'll also open up clinicians to do other things versus kind of educating on the fundamentals. Now they can actually be a better team in caring for the patient.
Host (47:54)
So with that, Andrew, thank you so much. You know, we ran over time. Thank you for your generosity and being here. It was a pleasure to meet you, and I look forward to meeting you in person at the World Medical Innovation Forum in a few weeks from now.
Andrew Trister
Yeah, likewise. I hope you have a great day. Thanks again.
