← Harvard Data Science Review Podcast
Harvard Data Science Review Podcast
Wrist Deep in Data: A Conversation With WHOOP Founder Will Ahmed
00:36:09
Loading summary
Transcript41 lines
- [00:01]
Liberty Vitter
Hello and welcome to the Harvard Data Science Review podcast. I'm Liberty Vitter, feature editor of the Harvard Data Science Review, joined by my co host and editor in chief, Shalime. Once thought of as a gadget of some futuristic fiction on the Jetsons, WHOOP is leading the charge in personal health from real time data, which revolutionizes how we eat, sleep, train and recover. This month we're diving into that process with whoop's founder and CEO Will Ahmed, who will guide us through the entire data journey from the moment it's captured on your wrist to the insightful metrics you see on your phone. Whether you're a health enthusiast, an aspiring entrepreneur, or just curious about how technology is reshaping our understanding of the human body, join us for an in depth call conversation on turning data into lasting positive change, potentially allowing us to live happier and much healthier lives.
- [01:05]
Shalime
Well, we all thank you so much for joining us. You have built WHOOP into a multi billion dollar business by turning invisible data into actionable insights. Can you walk us through the journey of data collected by WHOOP Sensor and from collections to the metrics we see on the app?
- [01:22]
Will Ahmed
Yeah, absolutely. Well, thanks for having me. I founded Whoop 12 years ago now and our mission is to unlock human performance and health span. Ultimately, we want to help people live healthier and longer lives. And we've developed the WHOOP sensor, which we're now on our fourth version of. And it collects the most data of any wearable on the human body. And if you think about the overall chain of collecting this data or the overall path, you know, first we have raw data that comes from our hardware sensor. So, you know, let's take photoplesmography for example, which is the technique for capturing heart rate. You know, we essentially have these various light forms that shine light underneath your skin, and the frequency with which those light waves return to a photodiode is the first input of data, so to speak. And that information then gets processed and analyzed. Now, if you're completely still, the frequency with which it is returning to a photodiode, well, that can be very directly attributed to what your heart rate is. Challenge is, if you're moving a lot, the sensor may be moving and your blood underneath your skin might be moving. So all of a sudden the, the idea that there's a perfect relationship there changes. And so we couple that data with raw accelerometry data, which is also measuring motion. And over the last decade, we've built all these algorithms around understanding the relationship of motion and photoplesmography to ultimately develop a very accurate heart rate sensor. And that heart rate data, depending on how hard it is to process, may live on the strap and be processed on the wearable itself. Or it might be processed on the phone, which is where we transfer the data to and then we take the raw heart rate data and then we start interpreting that into other scores. So I'm really just focused on one metric right now, but we do this for about 10 different metrics and one example of a score might be our strain metric. So we look at the amount of time that you would spend in an elevated heart rate zone versus a lower heart rate zone relative to your baseline. And the more time you spend an elevated heart rate zone, that would then contribute to a strain score, which shows how much exertion you've put on your body. Something like that strain score would feed into another set of algorithms and logic and coaching, which our large language model that sits inside the app would use to coach you. So, you know, the Punchline for the ND user, or as we call member, is Whoop recommending you should do a 12.5 strain today. And it should be a 30 minute run at moderate intensity. But the actual method of capturing what happened during that run or how we're talking to you obviously starts all the way at the raw sensor data level.
- [04:23]
Liberty Vitter
You know, one of the things that I've seen with Whoop that I think so interesting is this idea of actually changing people's behavior. So I mean, in your experience, have you seen how users have really changed their behavior based upon what Woops told them? And have you seen anything that really surprises you either by people's behavior or how they're changing their behavior?
- [04:44]
Will Ahmed
Yeah. So in the category of expected behavior change, we do see people listening to our coaching on exercise and strain relative to how recovered they are.
- [04:55]
Shalime
So.
- [04:55]
Will Ahmed
So if you are run down, WHOOP will tell you to rest. And in many ways, we were the first wearable to tell you to do nothing on a given day if that's what your body needs. And if your body is peaking more physically, so you have a high recovery score on whoop, WHOOP will recommend that you exercise or you work out. And so that sort of feedback loop has been very effective for behavior change in the app. And people who follow our recommendations are shown to be getting fitter. So we're quite proud of that. Also, under the umbrella of more expected changes, we do see people sleeping more consistently over time. You know, sleep is essentially this black box of your life for people who don't measure it. And so there's sort of an initial, I think, revelation of just how much is happening while you sleep when you start using whoops. And that awareness, I think, starts to lead to a sense of behavior change and action. We like to say you can only really manage what you measure. So if you're not measuring your sleep, it's kind of hard to know what changes you may or may not need to make to it under the umbrella of more unexpected behavior change. We've seen a profound shift in people's alcohol consumption on whoops. It's one of the biggest things that we've seen change. People drink way less alcohol after they've been on WHOOP for a number of months. And the reason for it is it shows up very visibly in your data when you have, you know, one, two, three drinks, depending on the person, it just cripples a lot of your numbers. So you'll see your sleep data really collapse. You'll see your recovery score very low. And for people who are looking to improve their performance or be more optimal in their daily lives, seeing that inspires behavior change.
- [06:50]
Liberty Vitter
So how do you tell that they're drinking less?
- [06:53]
Will Ahmed
Yeah, an enormous percentage of our members actually log in the Whoop Journal every day. And the Whoop Journal has over 100 different behaviors, lifestyle decisions, supplements, drugs, you know, things like hydration, alcohol consumption. All these different things you can track against alcohol happens to be one of the most popular things to track. And so we actually can see, okay, this person reported drinking this many, you know, nights a week in their first month on whoop, and, you know, six months later, now they're reporting this much alcohol. And so in general, we see, you know, a pretty meaningful decrease in alcohol consumption. We also get powerful testimonials from people. And I would say one theme has been around people getting help from WHOOP, realizing they had some kind of a problem, you know, going through a process of sobriety. And by the way, if you're someone who does have, you know, drugs, alcohol type problem when you get sober, it's amazing how your body responds. Like, all of your data looks so much better. And that that is actually inspiring. Right. For someone who's just trying to make what's probably a very difficult change in their life. So that was one of the more unexpected themes, so to speak, that we've seen with whoop.
- [08:11]
Shalime
Well, that's really good to know, but how do you assess whether there is a reporting bias? Right, because people tend to report more when they feel better, you know, that people probably is not doing well, or they didn't see much improvement, they just don't report or don't want to record their bad behaviors. Is there, is there some other independent way of you guys trying to assess the real impact of what you're doing? Obviously, which is tremendous, but just trying to see a big picture from the data science perspective?
- [08:41]
Will Ahmed
Yeah, it's a great question. We can also look at what their recoveries are over time as well. So if someone's average recovery has gotten higher and they're reporting alcohol less, those are the types of things that help validate that it's actually true. We also have a, we have a yes, no or no answer. So it's not like if I didn't fill the journal out today, WHOOP assumes then I didn't drink alcohol last night or I didn't do something, it just wouldn't count it, so to speak, towards the data set. And so those are the types of things that help ensure the data. One other thing that we, we do is we have the journal pop up first thing in the morning, often before someone has even seen their recovery score, their sleep data. So for example, like questions around mood or tiredness, those could be influenced by the data you're seeing. So Whoop says you got a red recovery and then it asks you how tired you are. Do you now feel more tired? Feel in quotes here because of what WHOOP already said. You know, we try to, we try to remove some of that bias. It's not perfect, but when you have it across an enormous data set, as you know, you get a lot of trends.
- [10:02]
Shalime
I think the first time I heard your podcast was on the Ultimate Human and I heard you talk about this heart rate of variability. And I was very excited about, because most people, they talk about tracking things, they talk about average or modes, but you were really talking about the variability. As you probably know, being a statistician, we love to talk about uncertainties. Variabilities tell us a little bit more about why the heart rate of variability is such an important one for predicting health and performance. And also what other kind of metric you're using is really about measuring variabilities instead of some averages.
- [10:40]
Will Ahmed
Yeah, when I was doing research on the human body, so this is now 12, 15 years ago, I was a student at Harvard undergrad, I stumbled upon heart rate variability. And there were some early studies on heart rate variability that showed really what a powerful predictive metric it was. So heart rate variability was being used by the strongest power lifters in the world as a metric to determine how much weight they should lift in a morning. It was being used by world class cyclists to determine how hard their workout should be. It was being used by the CIA for lie detection tests. It was being used by cardiologists to determine if former heart failure patients were going to have a heart attack. It was showing up in all these different papers as a very important and powerful statistic. Now, the problem with heart rate variability was that one, it's a very sensitive statistic, meaning it can change a lot over a short period of time, depending on what you're doing. And two, it required expensive medical equipment to collect. You know, like a $20,000 electrocardiogram machine in a hospital. So one of the fundamental questions in founding WHOOP was can we measure heart rate variability from the wrist continuously so we can manage some of the sensitivity and accurately so we can replace this otherwise expensive technology? And so that was really the first thing that we set out to do when we started whoop. Now, what is heart rate variability? Heart rate variability is this lens into your autonomic nervous system. And your autonomic nervous system really governs how your body functions. And the literal statistical measurement is it's the, the time between successive beats of the heart. So if your heart is beating at 60 beats per minute, it's not actually beating every second. This is a little bit counterintuitive, but it would be beating at say 1.2 seconds and then 0.8 seconds and then 0.7 seconds and then 1.3 seconds. And so the time between those successive beats is changing. And it turns out that's a good thing. The more variability, essentially, the more adaptive your autonomic nervous system is. Now, your autonomic nervous system consists of sympathetic and parasympathetic activity. So sympathetic is activation. So that's heart rate up, blood pressure up, respiration up. If you are stressed or about to exercise, or if you even think about something that's stressful, you may have a sympathetic response when you inhale. That's sympathetic, that makes your heart rate go up. Okay, Parasympathetic is all the opposite. So heart rate down, blood pressure down, respiration down. It's what helps you fall asleep when you exhale. That's parasympathetic. You want for every sympathetic for there to be a parasympathetic response. And that essentially shows that your body is balanced and in a relaxed or recovered state. And WHOOP measures heart rate variability all the time 24, 7. But in particular, we measure it during your sleep and during your Slow wave sleep so that we can calibrate over time how recovered you are. Because we compare that metric every night to your own personal baseline. And personal baselines can vary a fair amount. Again, because this is a fairly sensitive statistic. And the metric can also change a lot over the course of the day. If you just ate, if you're in a stressful conversation, versus on the flip side, if you just do a few simple breathing techniques where you exhale longer than you inhale, you can actually all of a sudden increase your heart rate variability. So it's a fascinating statistic. We could spend the whole time talking about it.
- [14:42]
Liberty Vitter
You're making me like breathe in and out. I found myself copying you as you were like breathing.
- [14:47]
Will Ahmed
Yeah, sure.
- [14:48]
Liberty Vitter
You know, I think one of the first things I ever heard about Woop was how it was really the first wearable to detect COVID19 patterns in users data. Could you like walk us through how you even figured that out? And did that sort of change how you thought about WHOOP in the future or did it have any real effect on how you guys thought about the company or product?
- [15:09]
Will Ahmed
Yeah, it's a great question. I think often in periods of crisis or unknown, an organization defaults to its value system. And for whoop, we were really founded with two core principles. One was do the research and the other was move really fast. And it turned out in building WHOOP that those two things often were in conflict with one another. Research institutions, especially good ones, tended to be pretty slow and then fast moving tech companies or otherwise didn't do the research. They didn't sweat the science. And so that became an important pair of principles. Now it's relevant because In January of 2020, a board member came to me and said, look, I think the rate at which this virus is growing in China, I think it's going to be everywhere. I think it might be a global pandemic actually. And he was an expert at growth, like he was someone who worked at Facebook and Uber and watched, you know, our knots all day. And so I said okay, well let's pretend for a second that were true. What would we want to do? Well, we'd want to be able to know who on WHOOP gets Covid and be able to measure what's happening to their body. Like we'd want to do the research. So we built COVID 19 tracking well before COVID 19 was a global pandemic. And then we released that in early March. And by the end of March of 2020, we had 2,000 people who had Tested positive for Covid. And if you go back in time to end of March 2020, that was one of the largest data sets in the world. So we had this very powerful data set, and we partnered with Cleveland Clinic and CQU to analyze the data. And we found a smoking gun, essentially for identifying Covid from a physiological data set. And it was a little surprising. If you look at someone who's healthy versus sick, the person who's sick will have an elevated resting heart rate, they'll have a decreased heart rate variability, they'll have disruptions in their sleep quality, and that's pretty consistent. What we found with COVID 19 is we saw all of those things, but we also saw an enormously elevated respiratory rate. Jumped off the page, increased about 20 or 30% off of baseline. Respiratory rate typically isn't something that changes all that much. So this was actually a fairly big finding because one, it was something that, for the most part, doesn't change. You have to be at altitude or be like, you know, smoking. And if you're not doing those two things, you likely have a lower respiratory tract infection, which, especially in the early days of COVID was. Was quite common. And so we had what was a smoking gun. And it even differentiated amongst the flu or a cold, because for the flu and cold, also, respiratory rate wasn't elevated. So we published this research in June of 2020, and it was peer reviewed shortly thereafter. And then professional sports, which is a world that we had worked in for a long time, they were all going back to play. And many of the athletes who wore whoop started to realize that they had Covid from their whoop data. And in particular, there was a golfer named Nick Watney, who, in the first PGA Tour event back, and the PGA Tour, if you remember, was the first sport to come back, he was cleared to play in a golf tournament because he tested negative for Covid. And the next day, he woke up with a red recovery on whoop and had this hugely elevated respiratory rate. And he essentially discovered that he had Covid from his whoop data. And he went back to the doctors and said, you need to retest me. And they said, no, you're cleared to play. You don't have Covid. He said, no, my whoop data said this. They're like, what is whoop? Blah, blah. Anyway, so he does another test, and sure enough, he has Covid. And so everyone was kind of amazed by that. But the PGA Tour then quickly put whoop on all the professional golfers, all the caddies, all the media members Everyone who is inside the bubble, so to speak, and look over the course of COVID we saw an enormous number of people realize they had Covid from their WHOOP data. And it was a really powerful, it was really powerful. I mean, we were quite proud to be able to help during that pandemic. And to your question, it did shift how we think about WHOOP as a company. For a long time, Whoop was under this umbrella of performance and fitness and sport. And we developed a health monitor during COVID which became one of the most popular features of the app where you could go in and see how your metrics were elevated or whether they were in line. And so now today, a lot of our focus is in and around just squarely health. We still have that fitness core to us, but we do a lot that's more squarely health.
- [20:12]
Shalime
That was really a fascinating story and you said something I think I'm sure many listeners would particularly interested in knowing more. You mentioned this issue about doing the research. Take time to do research, but on the other hand you need to move fast, right? It's always this kind of a contradiction, or at least a tension between time honored and timely. And so I think one thing you mentioned here essentially is a common problem. I know many data scientists like to be entrepreneur, but they're not entrepreneurs. You know, it's not just having technology can make you entrepreneur. On the other hand, I think there are lots of entrepreneurs would love to learn more about data science or AI, so they can lead with information, with knowledge, making informed decisions. You clearly have kind of strike a balance between both how do you do that and what are the challenges when you're trying to really take time to research but also moving as fast as possible.
- [21:14]
Will Ahmed
Well, I think moving with a real sense of urgency is core. I think removing as many barriers as possible. One advantage that WHOOP has versus research institutions now is we have a built in community that's really large. A lot of studies try to go out and recruit 50, 100 people for their study. And we can recruit people very fast through the app because we have this built in community. I'll give you an example. We saw that a cardiologist had done a very interesting research paper on heart rate variability in pregnancy. It was a three year study on 16 pregnant women. And the finding was that heart rate variability would decline steadily over the course of a woman's pregnancy, but then at the 33rd week it would inflect and start rising right up until the day that the woman delivered. And so the hypothesis was, is this 33rd week, an important moment in a normal woman's pregnancy. Now, we found this and we reached out to the women on our app and we asked if they wanted to enroll in a study. And in a week's time we had 2,000 women enroll. And then fast forward, call it six months. We had this enormous pregnancy data set, certainly the largest on heart rate variability in pregnancy that had ever been done. And so then we were able to look at this and sure enough, not only did we see this inflection at the 33rd week for normal pregnancies, we saw that 10% of the pregnancies delivered early, which is actually consistent with pregnancy in general. And then the question became, well, did their heart rate variabilities inflect at the 33rd week or was it seven weeks back from when they delivered? Because then it could be a predictive metric. Sure enough, it was seven weeks back. So now we've been able to publish peer reviewed research on using heart rate variability as a novel biomarker for early delivery. And all of that comes back to having this sort of energy for research and then being able to engage a community to participate in it.
- [23:29]
Liberty Vitter
That's such a cool story. I love that. It seems like you guys are sort of continually innovating, collecting new data, doing these studies. I mean, it's really sort of very cool what you're able to do.
- [23:41]
Will Ahmed
Thank you.
- [23:42]
Liberty Vitter
What is the new types of data that you want to collect? Like what, what is the future of this wearables data, the data that you currently have, and what data you might add? Is this like, I mean, is this going to become personalized medicine or I guess preventative medicine for everyone?
- [23:57]
Will Ahmed
Yeah, I think, I think the baseline will be preventative medicine. The challenge with the healthcare system today is so much of the costs are curative costs and in many cases it's a little too late to actually make a big impact. And if you can shift curative costs to being preventative costs, you can make the whole system function better and of course, have people be healthier. So a lot of how we think about it is around health outcomes. How are we driving health outcomes? How are we enabling someone when they put on WHOOP to have a definitive likelihood that they're now going to sleep more, or they're going to decrease their resting heart rate, or they're going to increase their VO2 max, or, or they're going to live 10 years longer? Like, what if we could prove that by wearing WHOOP that we're going to live longer or healthier Right. That's how we think about it. And from the standpoint of make the future feel like the future, I think it's inevitable that continuous wearable monitoring will be able to predict a heart attack before it happens or predict an illness or a disease state, you know, well before symptoms come on. And you're not going to see a doctor on some random year, you know, day of the year, where they take a bunch of vital signs that you could be measuring continuously and non invasively. You're going to go see a doctor like 30 minutes before something's wrong. And that should just make the system function better.
- [25:20]
Shalime
Do you get the users, because, you know, obviously people when they use, they see the benefit, they may want to measure this measure, that is, hey, can you help me to keep tracking of this and that? Do you have a mechanism for the users to suggest you new things to measure and have that happen that actually improved the WHOOP itself?
- [25:41]
Will Ahmed
We do. We've got a whole beta program where we work very closely with thousands of members on feature development. We've got another research effort called Whoop Labs where we're recruiting 50 to 100 people a day to come into our lab. They sign non disclosure agreements, they wear all these different sensors, and we're able to calibrate WHOOP against gold standard metrics for a very diverse data set. Something your audience might find interesting is that the diversity of your data, just human beings, is actually quite important. So because WHOOP was founded initially working with professional athletes, we very quickly found ourself working with a diverse data set because athletes in general have all sorts of different body shapes and different skin colors. And there's a big difference between measuring someone who has dark skin outdoors playing basketball versus someone who is indoors with light skin walking on a treadmill. Now what's the difference between those two data sets? Well, dark skin is actually harder to measure than light skin. No one really talks about that, but it's a fact. Thicker skin and darker skin are harder for photoplesmography.
- [27:09]
Shalime
I see.
- [27:10]
Will Ahmed
And then outdoors versus indoors, if you're outdoors, you have sunlight. Sunlight actually can start to affect the readings that you're getting. And then pick an activity like basketball. It's got all sorts of jagged movements. It's non periodic versus running, which is periodic. And so you saw the companies publishing heart rate data accuracy. That would be like, you know, 12 white men indoors on a treadmill, like that's super easy. That's not.
- [27:37]
Shalime
I see.
- [27:37]
Will Ahmed
That's not right. But if you take a huge data set very Wide data set. All of a sudden those algorithms will fall apart. In part because we started with athletes, we had to build this baseline or foundation algorithm that was pretty sophisticated. And then we just made that part of our principles where we always wanted to have a very diverse data set. So when we do a data collection, we'll have someone come in and we measure everything about their bodies, like how hairy they are, their skin color, obviously height and weight and gender and all these things. And. And that helps inform our data.
- [28:14]
Shalime
And that actually leads perfectly to the question I wanted to ask from day one, when I heard you talk about your company on the other podcast, is the data you have obviously is so close to people's life, right? You keep tracking of their body movement, lots of personal data. So there's an obvious question about the data privacy issue. And how do you build a trust with the users to give you all these data and how do you make sure that you do not run into these big data privacy leakage problem, which can cause tremendous problems for any company?
- [28:54]
Will Ahmed
It's a great question. I think data privacy is really core to what we do. At the end of the day, whoop sells a subscription. We sell a piece of hardware that comes with, you know, a subscription. And what we don't sell is you. Like, we're not in the business of selling data. We don't sell data to third parties, which is very different than what some of our competitors or other companies in the space may do. We use data to improve the product, which then in turn improves your experience. You know, we align with privacy laws like GDPR and ccpa. We try to follow security measures similar to a bank to protect your data. So we take data very seriously.
- [29:40]
Shalime
Yeah, I can just see enormous impact you have in terms of what you're doing. And protecting data privacy obviously is incredible for building the trust from the users. Before I turn to liberty to ask our magical one question, I have my own burning question, which is, you are Harvard educated and I want to know, does Harvard education is really useful? You can just say, no, it was useless. That's fine too. But I just want to get the sense of how much you feel you benefited from the Harvard education. But more broadly that, as you mentioned that you have been attracting lots of people to work for your companies. When you want to hire someone, what kind of skill sets you are looking for. I'm sure you have all kinds of criteria, and of course, I probably hope they don't need to go to Harvard. Right?
- [30:33]
Will Ahmed
That's right. I mean, look, going to Harvard was a great experience for me. I was a college athlete as well. So I was captain of Harvard squash team, which was a phenomenal experience and I think helped my education, especially for starting a company. You know, one advantage that athletes have sometimes over students is athletes every day have to face failure. You know, the thing that you failed at the game you lost, you have to talk about that and you have to work through that. And so there is a certain sort of sense of resilience, I think that comes from that as an experience. And I would encourage students to try to find that in their own life and not to do things because they think they're gonna do it well or do things because, you know, I'm gonna get a good grade in this. But to sort of put themselves in uncomfortable positions to grow, which I think is really what an education should make you comfortable doing and is entirely what's necessary to start a company or even join a fast growing company. Look, it's also worth saying I founded WHOOP out of the Harvard Innovation Lab. So I have to give Harvard a lot of credit in the sense that it gave me a place to get WHOOP off the ground. And for that I'm very grateful. I will say, as I've spent more time building this company, I would say overall, I'm skeptical of the educational background as sort of a core credential. I think the most important thing is that you're someone who's got a strong work ethic and you've got the talent to do the work. I think going to a Harvard or a top school might be a good proxy for talent, but you also need to fit into an organization culturally, and you need to have a certain work ethic to be successful. So those are things that I also like to lean on in the recruiting process. And by the way, we've hired a lot of great people out of Harvard, so we love Harvard. I think we're doing a lot of sophisticated things around data science and AI and computer science. And so for those listening who are interested in WHOOP as a career, I would encourage them to check out whoop.com careers and look at some of the roles that we have open.
- [32:47]
Shalime
Thank you.
- [32:48]
Liberty Vitter
Well, I don't think this was your goal in the conversation, but you definitely have a convert here. I'm very excited to order my WHOOP now.
- [32:54]
Will Ahmed
Good.
- [32:55]
Liberty Vitter
But I have sort of, I guess, a magic wand question that relates to a lot of what we've been talking about. So if you could wave your magic wand, what is the data that you would want to collect from Woop that there's no current technology to be able to do. Like what is your dream of what data you could collect?
- [33:14]
Will Ahmed
Well, I'll start with a simpler dream, which is that I think it's important to pull all the data in under one umbrella about your health. And so today you go to a doctor's office and you got those records that live somewhere. Maybe you go do a blood test and that lives somewhere. Maybe you take some supplements, maybe you're on a diet, maybe you wear a whoop. Maybe you work out with some specific type of product or you got some kind of a bike that's a connected fitness product. Like all this stuff is our data sets that don't live under one umbrella. And so we, one area that we're excited on and focused on is just pulling all that information under one umbrella to create essentially a home of your health. And in turn we should be able to connect a lot of dots that otherwise you've not been able to connect. And so I think that's where the health space will get a lot more sophisticated. I think, you know, without giving too much away about our future technology developments or some of the things we're working on right now, I don't think you need to look that much further than technology that you find in a doctor's office that's expensive. When I founded whoop, I said, okay, there's the electrocardiogram that measures heart rate variability. There's the PSG machine which is the gold standard for sleep monitoring. That's this big goofy machine. You have to go to a sleep lab, you have to put all these sensors all over your body. Okay, is there a way to do that? But non invasively. And there was a chest strap which is your sort of heart rate monitor for exercise, which is cumbersome and 40 year old technology. So I naively, I said, well, what if we could take those three technologies and put it in one small non invasive form factor. And I think you can imagine from going into a doctor's office what are the next versions of those? And those are the ones that we're working on.
- [35:13]
Shalime
Well, thank you, well again for this really, if not the most educational, but one of the most educational episode we have ever done. We wish you really the best. I think you're aspiration of putting everything together, you know, the preventive medicine. I think there's just a huge future there. Thank you so much.
- [35:32]
Will Ahmed
Thank you for having me.
- [35:37]
Liberty Vitter
Thank you for listening to the Harvard Data Science Review podcast to stay updated with all things HDSR. You can visit our website at HDSR, mitpress, mit.edu, or follow us on X and Instagram @the HDSR. A special thanks to our executive producer Rebecca McLeod, producers Tina Toby Mack, Arian Frank, Gavin Yang and Belle Reilly. If you liked this episode, please leave us a review on Spotify, Apple, or wherever you get your podcasts. This has been the Harvard Data Science Review. Everything Data Science and Data Science for Everyone.