← Lead With AI
Lead With AI
Your Medical Records Are Being Sold. Here's How Savva AI is Stopping It (with Stephen Rouse)
00:34:04
Your Medical Records Are Being Sold. Here's How Savva AI is Stopping It (with Stephen Rouse)
Loading summary
Transcript89 lines
- [00:00]
A
Today's episode challenges something we've all quietly accepted. That our health data lives in fragmented systems scattered across offices, rarely working for us in any meaningful way. But what if every doctor's visit, every symptom, every interaction could actually connect into a living, learning health story? Today's guest, Steven Rouse, is building exactly that. An AI powered approach to healthcare that doesn't just store your records, but tracks patterns over time and keeps your data where it belongs with you. In a world where privacy is increasingly compromised and data is often monetized, this conversation asks a bigger question. What happens when your health data finally starts working in your favor? Let's get into it. Welcome to lead with AI. I'm Dr. Tamara Nall. In each episode, we will take you behind the scenes with visionary leaders shaping the future of AI across public and private sectors. Join us as we explore groundbreaking projects and innovations that are transforming industries and making a real impact on people's lives. Let's dive in. So hi everyone and welcome back to the Lead with AI podcast. I'm your host, Tamara Nahl, or Dr. T, and I'm so glad to have you join us week after week after week because you listen to us and because you support us. We did hit number one in technology on Apple podcasts last year, as well as winning the W3 Gold Award for guests and interviews. And then hot from the press, we are an honoree for the Webby Awards. And so I'm just so honored to have all of that recognition. But it's because of you and because of great guests like the one we have today. Today we have Stephen Rouse, who is the co founder and head of growth for Sava AI. How are you, Stephen?
- [02:15]
B
Dr. T, thank you so much for having me. I'm so excited to be here with you.
- [02:19]
A
I am excited as well. Congratulations on all of your initial success with the MVP and the beta. We're going to get into all that, but yeah, thank you, thank you, thank you for being here. Now, our guests, of course, really want to know about Sava and what it does, but before we get into that, tell us about you as, as a founder, who are you at your core and at what point? Take us back to that time where you actually saw a gap in the market and knew that you had to solve it and thus Sava.
- [02:54]
B
Well, I've been querying my own, you know, I'm going to WebMD and ChatGPT and googling my own symptoms for various things that I've been dealing with over the course of my life for a forever, right. Since the since the dawn of the Internet and with the AI craze a couple years back, I was really surprised to find when you enter, you ask a question to ChatGPT and these modern AI systems that they come back with generic answers. That was the big problem, like, why do I have a headache? Oh, you might have brain cancer. Well, I might be hungover too. You don't need to go. I thought, wouldn't it be cool if, if, you know, you could connect your medical records and your wearables to these things in a safe way and have them respond with not generic answers, but, but answers that are rooted in your actual data. And about two years ago, my business partner Amit and I set off on this, this journey.
- [03:55]
A
That is amazing. So it all started with your own personal experience. Experience in your journey, of course.
- [04:01]
B
Yeah, it's. And then it's pretty funny, we were, we were having to explain to people the value need for this for the first year or two. And just within the last couple of months now we've got all these big tech companies that are asking for your medical records and your wearables data and getting into the space. And so now we've kind of explained to people what makes us different. And it's really exciting to be in the space right now. No doubt.
- [04:29]
A
Right, so you started it, so let's go there. What makes Sava different, Sava AI different than other companies?
- [04:36]
B
I mean, I think there's a fundamentally, all these companies who ask for your medical records, they say in a privacy policy, oh, it's safe, it's secure, it's private. Right. Or store your data. But it's de identified and for us, we don't collect any information in the first place. If you're de identifying my information, why are you asking for my email address and my, my phone number and my name?
- [05:01]
A
Yeah.
- [05:02]
B
And so we, we saw an opportunity, if you will, as a moat as the entrepreneurs out there want to talk about that privacy and security. Right. This is not consumer health data we're talking about here. This is medical data. This is the most private data that people have in their lives. And that for us, we didn't want anything to ever reach the cloud. So we have a on device model where when you download your medical records and you upload any information into the, into the app, it's actually landing on your phone and the app is getting bigger. Nothing is ever stored in the clouds as well as not having any, any accounts. We don't collect anything up front. So when you're done with the app, you delete it. There's nothing stuck on our servers. 23andMe sort of taught Americans what happens when one of these big companies stores your health data. All of a sudden it's being brokered off in bankruptcy court one day. We didn't want to play ball in that arena. We're really, truly trying to keep privacy and security as a paramount fundamental, not just a feature. It's in that functionality of the app. It's in, it's in the tech.
- [06:23]
A
Awesome. Awesome. Now walk us through that holy smokes moment. I mean obviously you've released the MVP. You have over 2200 beta users. Talk to us about that moment where one of your users experienced Sava AI and it changed everything for them.
- [06:42]
B
So I think it could go early on in the. Just between my co founder and I and various chronic conditions that we've been dealing with to. Over the course of my life I've had to go Into University of Utah's My Chart, UCSF's My Chart, you know, Stanford, MyChart. You had to log in and out each time. And over the course of my life, you know, and living down in Los Angeles for college, your. My health data is. And I'm not 65 years old, I'm in my mid-30s.
- [07:13]
A
Right.
- [07:14]
B
Just, just me I think, you know, five or six different portals just with the medical records data. And so we were able to put all that stuff in one place. Just the ability to see and parse through that data and, and not have to click in and out of portals. I think we saw it ourselves right. The first time the app was ever live on our phones. Wow. What a convenient thing to not have to deal with multiple my charts or different EHRs. You might have Cerner, you might have EClinical Works, all these different places. Your data. I have Athena data, I have Meditech data. It's just, it becomes a mess a lot. And so. Right. We've now been working with some of our. Most of our beta testers I'd say are consumer. Not consumer. Our patient advocates. A lot of patient advocates. Shout out patient advocates. You know who you are. Received a ton of feedback. Right. I mean people who are in their 60s who were like, wow. I was able to fetch medical records from a visit from a cardiologist 15 years ago when I was living on the other side of the country and I completely forgot about. And now it's, it's sitting there and I, and I, I know about it. Not only do they know about it but, but they're able to ask questions and get analysis from AI containing all that information in one place.
- [08:34]
A
Okay, well that's amazing. And you answered that for me because my question was going to be could it actually aggregate data from different siloed systems? And it sounds like it can, yeah.
- [08:47]
B
That's the whole point of the app. Right. Another differentiating factor since we're talking about it is you know, all these other systems you connect to your medical records via however they do it and your wearables and then they provide a blank chat screen as you'll see in the beta. I hope Dr. T is you. Actually we provide the raw data alongside the analysis. So whether you're looking for the after visit summary from that lab test you had done a couple months ago or you're asking questions, we combine all that data in one place and actually let you see the data, the rod coming out of all these different EHRs. It's pretty cool.
- [09:25]
A
That is cool. That is amazing. And so break it down for us because we're nosy, curious people. How does it work? Like yeah, we know AI trains, data, etc. Etc. But if we were to open up the hood and look at the brain of Saba, how does it work?
- [09:40]
B
Well, my, my co founder who probably listened to this will laugh. I mean I am ahead of growth and I'm the sales marketing go to market strategy of saba. I work with a brilliant technologist, his name's Amit Shah and he's amazing. And so I mean, what's truly under the hood? He'll laugh if I try, if I start pretending to know. But you know, I think very base in basic terms.
- [10:08]
A
Right.
- [10:08]
B
The on device storage is pretty simple. Right. But you get in the app and what we've done, we've leveraged something called Smart on fhir, which I don't know, give me a nod if you're familiar with the FIRE standard. No. So it's a, it's a health care specific API developed by the federal government.
- [10:27]
A
Okay.
- [10:28]
B
What we do is, you know, ChatGPT is connect, is has gone through Be well Health which goes through health information exchanges. Anthropic has gone through healthex, which goes through health information exchanges. Amit and I've gone directly to these EHR systems and to these providers directly.
- [10:51]
A
Okay.
- [10:52]
B
Access to this data. And so we leverage Smart on fhir, allow users to credentialize themselves in their own patient portals within our app and then locally store those records here on the device. Same thing whether you're rocking a whoop band or a, or a Fitbit or an Oura Ring or whatever it might be, collect that data as well, put that on the device and then enable people to use, whether it's a cloud based LLM like ChatGPT or Gemini or Groq, or use on device AI models like Gemini and Med and. No, sorry, like Med, Gemma and Quen and all those other ones that are on device and enable people to choose what they do. And when we send your data to the cloud, not our clouds, but if you're using grok, we encrypt that data and transit and rest. And so they never know who you are. And they're not, they're not storing the raw data, they're just getting little bits and pieces.
- [11:56]
A
Got it. Okay. All right. Amazing. That's good. Now, earlier you gave me an example of how, you know, you and Amit were kind of working in your early days and I would assume some users as well, where it was just amazing how it could aggregate all of this health data across these different systems. Tell us about a moment for you as the co founder where you experience something just as amazing or more amazing with Sava AI that it gave you goosebumps. You're like, oh my God. Like, we are. We have really changed the game here.
- [12:33]
B
So I had been wondering for years, you know, in college, it's a little cloudy, kind of forget what's going on. At least that was my story. I visited an emergency room down in Long beach and was in and out. And I'm okay talking about it. I had a varicocele. Really painful. Those listeners don't know what it is. It's painful in your private area.
- [12:58]
A
Okay.
- [12:59]
B
I was always, I came, you know, got home. I, you know, you're in college. What is the emergency room? A couple weeks later, was home before a break, my parents. What happened? I had no idea. I was like, I don't know. I was in the hospital. What was it? I don't, I don't know how years go by, right? I connected my, my data and, and all of a sudden, in my condition list in my, it said varicose. Like, oh my gosh. And it found it from 2011. Click into it, get some information of what actually happened on that day. And I had never, even the day after, I completely lost touch with this data point. Right. And so for me it was like, oh my gosh. Like, it's kind of scary, right? I couldn't believe that I could go back and find something I didn't even know existed in my medical chart.
- [13:54]
A
Oh my God, that's amazing.
- [13:56]
B
I think a month before that, if you had asked me the day before I had the app on my phone, I would have said, I don't know what it was. I had pain, I went to the emergency room. It sucked. Had a girlfriend at the time, come drive down, pick me up, went home kind of for lost track of it all. And so only imagine what happens with people that are managing serious chronic diseases out there to see. Wow, I forgot about that doctor and that urgent care visit when I was vacationing down in Florida that many years ago, whatever the case might be.
- [14:28]
A
Right. And so for me, yeah, you said that part. I was like, you probably went swimming in the ocean, a stingray got you or something. I don't know.
- [14:37]
B
It might have been. Honestly, I got to. I'll take a look at the raw data and see what it was.
- [14:42]
A
That is. Now that is really mind blowing. Wow, what a great example. Now. Now obviously you're working with a powerful tool here, a lot of people's private data. What are the ethical considerations that you have implemented and thought about as you were creating saba?
- [15:00]
B
I mean I think to answer that question, you gotta go back to something called the 21st Century ONC Cures act, which is a piece of legislation which I think has been around since the mid 2010s. Don't quote me on that. Some of your more serious readers are like 2012 or whatever. In 2022 there was a guidance given which allows for. It's pretty amazing from a population health perspective, allows consumers in this country to access their health records via third parties. Whether it's Sava or through GPT Health, whatever it might Google, whatever it is, Amazon, Microsoft, so really cool. Okay. From a. From an access perspective. Wow. I can. I actually own my health data. It's not stuck in some silo. And I don't have to go through my, my provider or for my insurance company. I can access this data via. Via third parties. Very cool. There's a bit in there that talks about how if you go through a third party, we as a third party in this example do not need to be. We're not required by law to be HIPAA compliant. I can do anything I want with your. And so what's happening and you know, just to. I already named names. I'm not going to do it again. But there's other companies out there.
- [16:22]
A
Uhhuh.
- [16:23]
B
Which we're talking about Google. How do. How does Google make its money? How does. Right. These guys are in the data game. Data being in 2026. And so we Again, going back to the first question, we don't put any of this data in the cloud ever. We store it locally on the device.
- [16:45]
A
Yeah.
- [16:47]
B
Privacy is part of the mechanism of the app. It's not part of a privacy policy. It's native to how the app functions. And from an ethical perspective, that's. That's exactly right. I don't want to play in that data game. We're not talking about, you know, consumer buying behavior and your, your geographics and all this other stuff that these companies collect. We're talking about some of the data points that people talk about with their significant others at the dinner table. It's going to kill them one day. This is not. This is not playing around. And so every one of these companies, a big shift into consumer health right now is awesome. But you got to be careful because what's actually happening to your data downstream, no one really knows. Right. And we've been lied to as consumers before. So we took the, the approach of saying, hey, man, we're not playing in that space. We're going to give people a tool that they can trust. And the problem with is that, you know, selling privacy and security is it's tough. Right? Because who's going to, you know, you take, take it, you know, straight out of the horse's mouth right now and say, okay, you know, this guy who owns this company is telling us it's safe, but why, why is he talking about safety and security? Too much, right? If I talk about too much, it's like, why? Starts people throwing up red flags. But I don't think the average consumer in this country understands that. Via the 21st century ONC Office of the National Coordinator in D.C. a governing body in D.C. because of this piece of legislation that was passed, these companies can do anything they want with your data, including selling it to an insurance company.
- [18:35]
A
Wow.
- [18:36]
B
And wow.
- [18:37]
A
Wow. Yeah.
- [18:39]
B
To cut you off, Dr. T. But the biggest concern. I'm rambling over her. But the biggest concern about that, right? HIPAA stands for health insurance Portability of the first three letters. Right. It's to protect ourselves from these guys so that, like, what if they find out, hey, Steve's Apple watch, he's a move. Or he uses my fitness pal and he logs 12 donuts and 12 beers a day like he's unhealthy. If an insurance company finds out with their big algorithms and all their money that I, that I'm pre diabetic before I know I'm pre diabetic, operation, right. All of a sudden your premium gets Gets. Gets, you know, through the roof or even worse, they decide that my family and I aren't even covered.
- [19:21]
A
Right, right.
- [19:22]
B
To keep this data out of the hands, it's not, you know, worried about some Russian hacker. It's wor. You know, I'm worried about getting in the hands of insurers and maybe potential employers. Maybe I'm going for a job interview one day and why would we hire Steve? Dr. T's healthy. I wanted to hire Dr. T. Why would you hire. He's, you know, an unhealthy individual. Right. You got to keep this data private. And so 21st century lnc. Cures act listeners read into that a little bit.
- [19:53]
A
Yeah, I'm definitely going to do it. I actually wrote it down right here. So when I look into it now, just I'm curious here. So if I. You saw the AI, what am I getting? Obviously, it's like I mentioned before, is combining aggregating all this data and it's a summary of, I guess my history of my conditions, et cetera. Does it also try to pre diagnose or is it just bringing it together and summarizing what am I getting?
- [20:15]
B
Yeah, I think that a big part of it too is, is you can go visit by visit over the course of your life. For me, back to the mid, like 2007 or 2006 on my app. And you could track data over time and biomarkers over time. You know, your A1C levels over the entire time, your weight, your. Your SpO2 max, all this stuff over time. Not just a white box that you can ask questions to, like a chat bot. Right. I think that's first and foremost the biggest use case of it. I think that, you know, if you're managing chronic disease especially it's a tool where you can, between visits, make more informed decisions on your health.
- [21:06]
A
Okay.
- [21:07]
B
That's the whole idea. We're not replacing doctors. We are not allowed by the FDA to make medical conclusions. We can't draw medical conclusions or give diagnosis, but we do go, hey, you know this blood test you got for this cough that you've been dealing with, you know, it's actually got some data in there that we think maybe you're leaning towards becoming a pre diabetic. You might want to check into that. Right. That it's looking at or maybe, you know, when you're going for. You start seeing a cardiologist at this place over here and on this place over here you have a rheumatologist and a way to have, okay, let's see this data side by side in a place where I can to go one one, a one glove fits all solution for all your medical data.
- [22:03]
A
Got it? Okay. I love that. Now talk to us about the big future. You know, five years from now, where should we expect to see Sava AI and how, what kind of impact and how will it change the world?
- [22:17]
B
So one of the coolest things we have going for us is that our direct connectivity to EHRs as opposed to going through health information, information exchanges and other vendors, not only is it more secure and private, but since we own that connection, we don't have to pay API calls. We actually own these fhir connections one by one to these institutions. And there's over 314,000 connected care sites in the country on my app today. I can list them. I'm going to show you a list. So when you get in there, you search by city, you get searched by doc. You can do a bunch of stuff and you'll find you'd be able to get in and see that data. So very cool from, from that. So we, we could make money on $10 a year. The app costs $10 a year. Right. Part of the reason we're making it so cheap is that we have massive plans for outside the US and in rural United States as well. My father here in Sonoma, right here in the Bay Area, his cardiologist is not on an EHR system. Oh, even in northern California. Right. Tech capital, it's still not on an EHR system. And it's not that, you know, it's, it's not that unique. I think 35% of Americans don't have electronic health records. And so what we allow people to do and we enable people, whether you're living in rural US or anywhere else in the around the world, you can connect your, your data to the app and how have it stored locally just by uploading pictures and documents into the app.
- [23:52]
A
Wow.
- [23:54]
B
Within five years, 100 million people around Western Europe and the United States I can get via EHR and I can, I can have them be on the phone in the same time period. My goal is to have a billion people around the world in underserved regions, in places like Dar es Salaam, Tanzania, where I have been hospitalized before happen long story for another time. Basically fell off Mount Kilimanjaro right when I had first met my wife. Not, not a good look. Anyways, in places like that, where modern EHRs at $1 billion a build and all these big EHR systems, they're not coming to, to those places in my son's lifetime, right? Those, those individuals, those patients show up with a, a notebook just like our grandparents did, right? You show with a big binder, you staple your stuff in. And so five years from now, a billion people will have digital health records around the world because of Sava, because of Saba and giving them the first taste of digital health and tracking that information all in one place.
- [25:05]
A
Yeah. Wow, that's. That's absolutely amazing. So, I mean, you've convinced me for sure. So if our listeners want to be connected, roll up their sleeves, use Sava this week, as in today, after listening to this episode, what's the best way to do that?
- [25:21]
B
So it's. It's www.sava sava.AI. you go to our website and sign up for the beta, right? Then you'll. It's on test flight only. So for Android users, you got to wait a little while. But all iPhone users can get on that way. And then very soon here at the end of April, we'll be live on the app store on iOS. And so check it out. There's a free version. And again, the premium is only 9.99 a year.
- [25:52]
A
That's amazing.
- [25:54]
B
To use the full app.
- [25:58]
A
That is absolutely amazing. So now we're going to move to something that I call from one genius to another, where a previous guest has a question for you, Steven, and that question is, if AI ever became conscious, how would we even know?
- [26:13]
B
Who says it's not already conscious?
- [26:15]
A
Okay. I like the take,
- [26:19]
B
you know, or it'd probably start doing therapy like the rest of us and blaming its training data or something like that.
- [26:27]
A
Who knows, huh? Well, a lot of people are using it for therapy because there's. It's. Sometimes it's so hard to get an appointment, etc. So what you're saying is we already there.
- [26:40]
B
How could we know? I mean, some of the stuff I see, especially with the images and videos, I get caught all the time. I like, I'm like, I'll show my wife something. Did you see? She goes, that's not real. I'm like, oh, my God. So cooked, right?
- [26:56]
A
Who knows?
- [26:57]
B
You know, who knows? It's getting so good. I think that with the dawn of quantum computing, right? If someone could figure that out. I mean, I know it's around now if you, like, blast with liquid nitrogen or whatever, but if someone could get room temperature quantum computing done in our lifetimes, I think it's certainly a possibility even within the next 10 years work so cooked. I mean, it's an insane technology and I think that we really need to call on our leadership to make sure this is governed appropriately, this new shift.
- [27:34]
A
But, yeah, thing I would say about that, because I do get asked that often with AI in terms of regulation, governance, policy, et cetera, when it comes to regulation. Given that I've been in the federal government now for almost two decades, I get to a point where I'm thinking, I like regulation, when done correctly is amazing. But I often find that by the time it actually passes both sides of the aisle and you get everybody on board, it's 180 degrees different than its initial pure intent. And so, I don't know, I think there's something that needs to be done around that and that we always go back to, what is the initial intent? What are we trying to promote, what are we trying to avoid so that it doesn't, doesn't be like, okay, I'll pass this if you do this for me, you know, Amen.
- [28:26]
B
I mean, I think I couldn't agree more. I don't have a brain big enough to figure out how that works. Relying on the government. Let's rely on some of these founders to, to do the right thing as, as these things are being built, right?
- [28:43]
A
Bingo. Bingo. It starts with you as the founder and the developers to make sure that's the right place. Because I've even been asked, oh, well, what about AI for bad? I'm like, sure, there could be AI for bad. It depends on the intent of the developer. Just like, I don't know what the number is, but forever, for every a million people don't know what the number is, there's a serial killer. Okay, well, the chances are the same with AI you can still have that developer. And that's why it's up to us to make sure it's unbiased and that it's ethical and that it still has this pure intent to help or whatever.
- [29:18]
B
So people need to just wake up and look at some of these companies and some of the people in charge of them, and would you trust them with certain data sets? And I mean, I just, I could look at an interview with some of these leaders and like, I don't know, that's not my guy. You know, I, I think just, you know, let's use our common sense here.
- [29:40]
A
And that is key. As my, my, as my grandma used to say, God bless her soul, common sense ain't so common.
- [29:49]
B
Yeah, a lot of people try to outsmart their common sense, and it's like,
- [29:54]
A
go with that original gut, that initial gut is there For a reason. So that's amazing. All right, Stephen, we're going to move into our rapid fire questions. I'm going to ask you four questions.
- [30:06]
B
Haven't been rapid fire so far. Oh, God, no.
- [30:10]
A
I've been pretty easy on you.
- [30:12]
B
Okay.
- [30:13]
A
Okay. Take your sip of water. I'm going to ask you four questions and you give me the first response that comes to mind, starting with, what is the most overrated AI tech trend?
- [30:26]
B
Man, this is tough, Dr. T. The most. I know you want quick responses. The most overrated AI tech trend. AI and the wearables themselves.
- [30:41]
A
Overrated. Okay. Oh, AI in the wearables, like by themselves?
- [30:48]
B
Yeah.
- [30:48]
A
Okay. All right, I get that. That, that leaves room for Saba. Okay, I get it. I like what you did there. Okay, let's see how you answer this one. What about the most underrated AI or tech trend?
- [31:03]
B
I think what we briefly touched on with this quantum computing and the future, I think we might be living in the fifth industrial revolution right now. And the sixth will be the dawn of quantum computing. And I think it's not being talked about enough. And it's an underrated point. Part of. Of what machine learning and AI is today.
- [31:27]
A
Got it. Okay, got it. What's a book we should all read?
- [31:31]
B
Oh, my gosh, read. You know, I just read a good one. Never split the difference. You ever see? You ever read this one?
- [31:40]
A
I've heard of it, but I have not read it.
- [31:42]
B
The the Art of Negotiation. This is a good book. I really liked it. I really like that book.
- [31:48]
A
Got it. All right, now. Scare us. Wow us. What is your biggest, boldest prediction? It could be about anything. Business, AI, personal. Like wow us with your biggest, boldest prediction.
- [32:04]
B
My biggest, boldest prediction that within 10 years, room temperature, quantum computing has completely shifted the way we think about every sector of business in the world, from agriculture to big tech to banking to government, across the board.
- [32:28]
A
Awesome. I love it. Stephen, thank you so much for being here. I really enjoyed this conversation. All the good gold nuggets that you've given us about your vision, how you even thought about it, you would admit Sava AI, if we want to get in contact with you, if we want to know all of the social medias for Sava AI, tell us all of the those so that we can make sure that we connect with you across the spectrum.
- [32:54]
B
We're etsava on all the social media platforms. I am very active on LinkedIn. Please give me a call. I would love to meet. I meet with people all day long every day. And honestly, if you're ever in the small town of Sonoma, California, if you go down to the Plaza and just scream my name, I was born and raised in this little town and I will come down and we'll hang out. That especially goes for you, Dr. T. You got to come out and see us in wine country one day. Absolutely.
- [33:21]
A
I will do that. And then we're going to really find out that that headache was really a hangover. Oh, my goodness. Absolutely. Again, thank you so much for Stephen. I really enjoy the conversation. And and to everyone who's listening, thank you for tuning in again. And remember, until next time to lead with AI Take care. Bye bye. Thanks for tuning in to lead with AI I'll see you next time as we continue exploring the cutting edge innovations shaping AI across the public and private sectors. Until then, keep leading with AI.