Prime Venture Partners Podcast

He is building the Future of Healthcare for the World from India - Dozee's Cofounder Gaurav Parchani

• Prime Venture Partners: Early Stage VC Fund • Season 1 • Episode 152

Share episode

In this enlightening episode of the Prime Venture Partners Podcast, we delve into the transformative world of AI-powered healthcare monitoring with Gaurav Parchani, co-founder & CTO of Dozee. 

As India’s 1st AI-based Contactless Remote Patient Monitoring (RPM) & Early Warning System (EWS), Dozee is FDA cleared, has monitored 250 Billion+ Heart Beats and saved 14 Million Nursing Hours! 

Dozee tracks vital parameters such as the Heart Rate, Respiratory Rate, Blood Pressure, Blood Oxygen Saturation, and Skin Temperature with clinical grade accuracy, and tracks sleep quality while flagging indicators of sleep apnea.

In this podcast episode we spoke about the below topics, dive in:

00:00 - Revolutionizing Healthcare With AI Monitoring
07:50 - Reimagining Patient Monitoring With Dozee
11:13 - Transforming Patient Monitoring With AI
21:42 - Advancing Healthcare Monitoring With AI
28:13 - Doctor Skepticism vs Real World Evidence
40:44 - World's 1st Non-Contact Blood Pressure Invention

Enjoyed the podcast? Please consider leaving a review on Apple Podcasts and subscribe wherever you are listening to this.

Follow Prime Venture Partners:

LinkedIn: https://www.linkedin.com/company/primevp/
Twitter: https://twitter.com/Primevp_in

This podcast is for you. Do let us know what you like about the podcast, what you don't like, the guests you'd like to have on the podcast and the topics you'd like us to cover in future episodes.

Please share your feedback here: https://primevp.in/podcastfeedback

Gaurav Parchani:

A normal person would think okay, so you're fine, and then you slowly deter it, and then you further deter it, and then you go to ICU. Right, it doesn't happen that way. You'll start breathing faster and heavier Heart rate, respiration rate, blood pressure and pulse, oxygen saturation.

Sanjay Swamy:

Today we are 2 million nurses short in India, right, what's possible now? That wasn't possible earlier.

Gaurav Parchani:

We got the data to the nurse instead of the other way around, and healthcare is definitely due for a big tech revolution. I would say so that we can identify close to 91 to 92% of patients eight hours in advance. So Dozee is the world's first contactless remote patient monitoring solution, along with an AI-powered early warning system, because every patient's body is different, everybody's baseline is different. A sensor under my mattress capturing all the vibration can actually capture that dimension as well. We are present in UAE, we are present in Africa and, more importantly, we are present in the United States as well. Right, but non-contact. We are the first one in the world to do.

Sanjay Swamy:

All right, I have my good friend Gaurav Prachani here, one of the co-founders of Dozee and notably, according to the other founder, the smarter of the two. So the audience will vote after this. But, gaurav, it's been a great journey, you know, working with you guys and, over the past four years, really interacting closely. So welcome to our podcast. is sort of at the forefront of artificial intelligence and it's used in, used in uh in the healthcare field, particularly around patient monitoring and um, early detection of uh, uh and hopefully saving a lot of lives. So, uh, welcome to our show. We'd love for you to share a little bit about your background to start with uh, and then you know how you came across this idea for and you know then we can dive into more details.

Gaurav Parchani:

Sure. Hi everyone. As you already introduced me, my name is Gaurav. I come from Indore, my father works for Department of Atomic Energy and in fact I grew up around construction sites and these are not your everyday construction sites. Uh, industry, which is asia's largest particle accelerator, was a few kilometers from where I stayed, so I've seen it being built and that's how I think I got curious about engineering, mathematics, problem solving, uh, and that's how I started kind of looking into a lot of and was always interested into mathematics.

Gaurav Parchani:

I got selected into iit indore, again close to my home. So everything in indore so far, first batch of iit, indore, mechanical engineering, and there I got a lot more into automotive engineering, like got a lot more interested into that, particularly a lot of simulations around automotive industry. So your computational fluid dynamics how does the fluid or airflow around a vehicle affects its drag or crash dynamics, how, when a vehicle crashes, what happens exactly? Right, and how do you make safe vehicles. And right after college I joined this company, an American company based out of Bangalore. They have a big development center in Bangalore where I met Mudit. Again, our team was simulation engineering plus a lot of sensors, developing prototypes so that we can make cars go faster, be safer as well, and also the product lifecycle management is lesser as well. Right, like the lesser cycle, so you can develop cars faster and cheaper as well.

Gaurav Parchani:

We worked for a few years over there, but then after some time, we always felt that whatever we were doing and this I'm speaking more for myself now whatever I was doing was amazing, as in these were amazing problems to solve. An always requires a big, a good problem to solve. More complex the puzzle. More fun. It is uh. But at the other end of the spectrum is also that it should also create a real life difference as well, right? So yes, in this case there was a difference. We were interacting with a lot of clients ferrari, mercedes, porsche, toyota all of these were clients of the company and we were interacting with these. We got to see what was going to the market and how our softwares and our simulation solutions were affecting it. But it's not that large of a difference in terms of impact to society.

Gaurav Parchani:

Mudit came back from a trip, I think from Germany, and this was the time when an unfortunate event occurred in his family, right, and we started thinking that it's been like, if you look at the last two decades, every other industry has completely been revolutionized by data and by tech. Right, I know the LLMs are the craze nowadays, but, but even before that, we started almost nine years ago, and at that point of time, we analyzed, two decades ago, imagine how did you get your advertisements? Maybe radio, tv, newspaper, roadside hoardings? That's it. Right Now, each one of us gets personalized ads served to us based on what we click. The industry has completely changed Healthcare. However, however, we follow the same protocols and that were, decades ago, set up. Uh, whether it, whether it is for care, whether it's for patient safety, whether it's for monitoring patients, keeping vigilance on them, right, uh, and this is something that, uh, both of us uh that this is high time now that somebody should work in this direction, and healthcare is definitely due for a big tech revolution. I would say so.

Sanjay Swamy:

One of the reasons and this is we are talking, you guys are like 23, 25, that 23 around.

Gaurav Parchani:

Yes, I think I was 23, he was 24.

Sanjay Swamy:

I was just trying to reflect on. What I was thinking of when I was 23 was not about why healthcare is lag lacked for sure, so that's pretty cool so, uh, and this is this is where, uh, we started analyzing why is that so?

Gaurav Parchani:

right, like, uh, why is healthcare?

Gaurav Parchani:

It requires much more patience. It requires time. Uh, you have to. These are not only software solutions that you would fit in. Right like, you have to be ready for whatever is required.

Gaurav Parchani:

If we didn't want to build the hardware in the first go itself, right, uh, but we figured out. There's no other hardware that can give us data on patients, right, like. The idea was, yes, e-commerce companies nowadays can figure out and even that point of time when you're going to order your next toothbrush right, uh, but why can't we figure out, uh, when a person is going to order your next toothbrush? Right, but why can't we figure out when a person is going to crash next hour, next day, next week, next month? Right, that sort of risk modeling on a continuous basis did not exist, and for us, the major reason for this was was lack of data. I think we have amazing engineers throughout the world that can forecast almost everything. You just need to have the data and a good amount of data for it.

Gaurav Parchani:

Right, and here the fundamental problem was there's no data continuously available for patients, whether they are at home or whether they are at hospitals or wherever they are, and healthcare particularly, being slow, primarily because it's a very highly regulated industry, in my opinion. I think I would say it is the second highest regulated industry after space tech, and rightly so. You're dealing with patient lives, patient safety but it can be difficult for somebody without a lot of highest regulated industry after space tech uh, and rightly so, you're dealing with, uh, patient lives, patient safety but it can be difficult for somebody without a lot of backing, funding, experience, uh, to start something and survive in this particular industry. Uh, I think, with a lot of, I think, a little bit of luck, with a lot of hard work and amazing partners such as yourself, I think we've been able to beat the odds there.

Sanjay Swamy:

But, yeah, it's been a very rewarding journey and get to the start line right. I mean, the journey is ahead of us.

Gaurav Parchani:

in that sense, yeah, I think nine years now, but I think it still feels like we've just begun right. So I think it's still there's a long, long way to go, but yes.

Sanjay Swamy:

So awesome. I think that's a great, uh you know introduction to your background and uh, you know the why uh thing. So tell us a little bit about what is and maybe you know just a couple of uh, like a twitter style, uh, the original twitter style responses for for getting our audience little oriented. Uh, what's the problem? What is ? And you know where have you all reached today in this journey?

Gaurav Parchani:

So is the world's first contactless remote patient monitoring solution, along with an AI-powered early warning system. What's the problem that we're essentially tackling?

Sanjay Swamy:

Can you just break that down? That's like a mouthful in itself.

Gaurav Parchani:

Okay, Right for people to be able to understand each of those yes, so is world's first uh remote uh contactless remote patient monitoring solution and an ai powered early warning system. Uh monitoring traditionally happens with a lot of these contact-based probe-based systems. If, unfortunately, somebody's been in ICU, they would know about it. You have ECG for getting your heart rate or your rhythm. You have a cuff-based blood pressure for getting your blood pressure. You have a nasal calunia put into your nose to capture your respiratory cycles, right. You have an oxygen saturation probe that's put on your finger and so on. There's so many wires and everything put on. So for current standard of care, if you have to be monitored in a hospital, you have to be put on basic minimum of these four to five probes on your body. We can go more invasive if the patient is more risky in ICU, but outside the ICU the patients are not that risky. So you have to put all of these on. And these are all bedside monitoring. So there's a screen there where the data remains next to the bedside. So the nurse or the healthcare professional or the doctor has to come to the patient, to the bedside to actually look at the data, right.

Gaurav Parchani:

There are major reasons where this doesn't work. First, outside the ICU, patients are ambulatory. What I mean by that is patients get up, move around, right. They may go to the washroom, come back, or they may be scheduled for an x-ray. They'll go for an x-ray and come back right, and if we ask the nurse again and again to put these probes back on the patient, it's going to be an operational nightmare. Today we are 2 million nurses shot in India, right? Imagine now I'm asking you more nurses for operationalizing monitoring particularly.

Gaurav Parchani:

Second problem directly related to nurses data is where on a screen where the nurse has to go to the data, right, and this is exactly where, outside the ICU with because of how WHO prescribes it, we need to have one is to four, nurse to patient ratio. But in the best of hospitals, you will see one is to six, one is to eight, and as you keep going down to tier two towns, as you keep going to public hospitals, you will see one is to six, one is to eight, and as you keep going down to tier two towns, as you keep going to public hospitals, you will see one is to 10, one is to 15. Personally, I've seen one is to 30 as well in the country. Now imagine being that nurse and looking at 15 screens at one time. Yes, you could do it in an icu, because icu is one is to one, so you are associated with one patient for six hours so you can take care of that patient. But imagine being associated with 20 patients, right, and looking at 20 different screens. Not possible.

Gaurav Parchani:

So that's why, at Dozee, we decided can we rethink monitoring? And in that can we get rid of all the wires? And wherever we cannot, we'll make it wireless. But just let's get rid of all the wires and in fact let's get rid of contact itself. Right, can we make monitoring as easy as being part of the furniture or something that is very passive in nature? Usually when people are making products, they think of an active engagement and so and so forth.

Gaurav Parchani:

Our thought was completely reverse. We want to automatically collect data. We want to be as passive as possible. The patient should not even notice, right, their experience should be so good. And this is how we solved by making the sensors contactless.

Gaurav Parchani:

And I'll explain about the sensors. We get data continuously. All the patient has to do is lie down on the bed, no wires to be connected. If they go to the washroom, come back those 10 minutes later or recording automatically starts again. Second, we got the data to the nurse instead of the other. Right, so we connected all of this to the cloud and gave nurses access on their nursing stations on their smartphones. Even doctors, while they are outside at their home, or maybe in their procedures or in their OPDs, they could actually get to see. Okay, I have 10 patients admitted. These two are at very high risk. There's an alert on one of them, right, so we got data out from the bedside from so we got data out from the bedside from all of the patients to the healthcare professionals. These two are the major paradigm shifts, but this is where the major, this is where an issue that we can generate out of this Imagine again being that nurse, just to back up.

Sanjay Swamy:

So you said two big things. One is, instead of sensors being fitted on the body of the patient, the ideal situation is there is no sensor fitted to the patient and that's what makes it contactless. Yes, in a few situations you may still need to have contact, but it still doesn't need wires. It'll be wireless, but it might still be a patch or something like that is installed absolutely attached to the patient absolutely so.

Gaurav Parchani:

We get three vitals, contactless uh your respiration rate, heart rate and non-contact blood pressure and for oxygen saturation and temperature and an ECG rhythm. We have three other separate modules that will be wireless completely without compromising patient experience and compromising patient safety as well.

Sanjay Swamy:

And all the data comes to the nurse station. So if I'm a nurse, I've got this. You know cockpit, so to speak, or you know control center, so to speak, and I'm monitoring everyone, but I may not even be physically in the hospital, right? So a doctor could be monitoring Absolutely.

Gaurav Parchani:

This has given rise to and this is something that I didn't even think of when we started dosing nine years ago but this has given rise to command centers now completely based on this Right nine years ago. But this has given rise to command centers, now completely based on this right and command centers remotely managing multiple hospitals in one shot right. So there is a multi-layer escalation system where the nurse gets some data she has to take some action on. If they miss it, then a hospital level RRT rapid response team gets it, and if they miss it, then there's a command center level RRT rapid response team that essentially get to it. So RRT rapid response team that essentially get to it. So your your rest assured that your relatives or whoever is in the hospital right is well monitored, is being under constant surveillance and is getting the best care that they deserve so for us cricket fans is like having a third empire sitting, yes, somewhere else, not at the stadium and making decisions yes.

Gaurav Parchani:

So now imagine again being that nurse.

Gaurav Parchani:

Yesterday you had four vital readings per patient, right? Why? Because of all the reasons that I spoke about, right, monitoring is not possible outside ICU. So today, the current standard of care, and just to give you some numbers, in every country, including India, close to 90 to 95% of hospital beds are non-ICU beds. So we're talking about 2 million hospital beds in India and 1.875 million beds of them are non-monitored outside the ICU. Now, on all of them, such monitoring is not possible or continuous monitoring is not possible.

Gaurav Parchani:

So what we end up doing and this is what I meant by last two decades or three decades we've been doing this spot checks. So nurse has a round schedule every four hours to every six hours, depending on the hospital protocol, depending on the patient condition, they'll go next to the patient. Take all the readings, five readings. Put it on a chart paper. When the doctor comes around they'll get to see four dots connected with lines. Is that nearly enough to get the trends or the, the picture of the patient right, like why get four picture, four images of the patient when you get a full high definition video right? And this is what leads to the need of monitoring. But imagine again being that nurse. You, yesterday till yesterday, you had four values per patient per day. Today you have almost for hundreds of patients, a value a minute, right? So much of data it's. It can be overwhelming.

Gaurav Parchani:

And this is where ai comes in because, uh, in order to make risk stratification, in order to triage patients, who is at higher risk, who needs attention first, right, uh, where you may need urgent care, that sort of risk stratification is something that ai does and that is what we call early warning system. And there are multiple types of early warning systems around the world, but they are all statistical in nature. They all have to be hand calculated because they've made easy, uh, for somebody to collect these vitals quickly, calculate it and, on the back of their envelope, and be able to kind of take decisions. Basis that. But with ai, we, we starting. There we're, we're exactly mimicking what early warning scores and systems have been doing across the world, and nhs and uk being one of the leaders that they've made full protocols around the system and it's a standard process that they follow.

Gaurav Parchani:

But this is where the future lies. We can go way beyond that. Why be limited by human capacity to consume five numbers or four numbers right, when we have a stream of vibration data coming from the sensor which is at least 250 to 500 samples a second A second, not a minute. So we are seeing much more dimension data and much, many more dimensions than a human is getting to see, and this is where we have the opportunity to not just replace what humans have been doing, but actually go one step beyond as well, and that's where the future lies of early warning systems, and that's what I'm particularly very, very excited about.

Sanjay Swamy:

Wonderful. So what you're saying is initially, you know, there was a lot of manual effort in getting the data. Now you've automated the ability to get the data and by making it contactless, it's easier for people to actually do it, to capture the data, etc. Bring it. But now you've created a new problem because there's too much data and the nurse that was struggling to get the data, et cetera. Bring it. But now you've created a new problem because there's too much data and the nurse that was struggling to get the minimum amount of data is now suddenly being overwhelmed with a lot of data. And that's where the analysis that you're doing and extracting the alerts plays a big role here. So all this is to try to do what they were already doing, but making it possible to do it for a larger group of people and more consistently. And now you're saying there's there's something beyond what they were currently doing. So tell us more about that um, so yeah uh, what's possible now?

Sanjay Swamy:

that wasn't possible earlier? Yeah, I guess that's the question yeah, absolutely so, it's.

Gaurav Parchani:

It's almost essentially when a person decompensates right and when a person is basically going through a cycle of health deterioration, whether you're coming from whichever comorbidity, let's say if it's a liver patient or whether it's a kidney patient or whether it's whatever neurological patient is. Whenever a code blue is announced in a hospital and I'll define a code blue usually, code blues essentially mean emergency events where a patient requires emergency care and they're either shifted to an icu in an emergency or, unfortunately, the patient passes away then and there itself, and this is where you register a code blue. Every hospital has a procedure to do that because it's required by regulations and compliances where they essentially record and they're essentially they. They need to have processes. When this happens, who will get come and administer care? How will you triage? How will you diagnose? How will you give, give patient better care?

Gaurav Parchani:

Now, what we've seen in research right, there are signs uh which deteriorate at least four to eight hours in advance. At least it can be even more as well for some people right, where you would essentially see there is a clear deterioration in cardiopulmonary uh systems, so insufficiency of cardiopulmonary systems. What that means and if I translate it to normal English. It's essentially no matter which comorbidity you're coming from, whether you're coming from liver, whether you're coming from neurological disorders or cardiovascular health or whatever. Finally, at the end of the day, the code blue happens when you crash or when a person crash. It only crashes when either your pulmonary system is crashing, so your lungs are either filled with fluids and you cannot breathe properly or you're going breathless, or it's essentially that something happens to your heart and your heart stops at some point of time. Right, so it's either the heart or the lungs, so cardiopulmonary insufficiency insufficiency meaning they're not able to perform the system or the the, the output that they were supposed to give in terms of blood or in terms of the oxygen that you need to get. You're not getting that. Now, there are four vital signs heart rate, respiration rate, blood pressure and pulse oxygen saturation. They have a clear relation to this particular event that occurs. This has been going on for so many decades, but it's so unfortunate that we don't have majority of data around these processes on how the cycles of decomposition happens. Right, so a normal person would think, okay, so you're fine, and then you slowly deter it, and then you further deter it and then you go to ICU? Right, it doesn't happen that way.

Gaurav Parchani:

There are cycles of these compensatory mechanisms that the body induces itself. So imagine in order for you to function, you require a lot of oxygen. Right, that oxygen is going to every cell in your body. Who is taking that oxygen? To your every cell in your body? The hemoglobin in the blood. Where is the blood getting it? Because you're inhaling it and your lungs are actually transferring the oxygen there.

Gaurav Parchani:

Now, if you require more oxygen, right, and it's not reaching your peripheries, or it's not reaching your other part, other body parts or other organs, for example, right, the first compensatory mechanism of the body is to increase your respiratory rate. It's the easiest thing the body will do, and so you'll start breathing faster and heavier. That's the first mechanism, and then you'll start feeling fine, a little bit right, but this is not the cure. It's not a cure, right, again, it's. It's going to go, go out of hand. And then what essentially happens? When it goes out of hand, you will. The other mechanism that the body has is the blood. So if, if I'm not getting enough air in the blood, I'll send more blood, that will reach, uh, the body as in faster, uh, so your heart rate would go up, or your blood pressure would go up as well, right, uh, and then finally, everything, nothing works out. Your oxygen saturation will drop, right, and these three or four vitals keep going in combination, up and down, and up and down and finally, after a cyclical nature, there comes a point where you say that, okay, this person is not able to compensate at all, and then they crash. That is where you announce a code blue.

Gaurav Parchani:

So these abnormalities in vital signs can actually be picked up with proactive alerts when you set up continuous monitoring around these vitals. Now, this is the statistical way to do that, right, but these ways and this is quite effective, by the way. So in fact, we ourselves, using these techniques, have proven in clinical studies that we can identify close to 91 to 92% of patients eight hours in advance by these abnormalities in these vital signs. But these vital signs are essentially now they're effective 92, as I said. The dark part behind that is that they're very, very sensitive, which is amazing, that they are capturing all patients, but they are not very, very specific. What that essentially means is that, yes, I'll get 10 alerts for a patient who has to go to ICU, right, but I might also get one or two or three alerts for a patient who is actually doing okay or is doing moderate and then is recovering faster and then going home. So it's actually increasing a lot of alarm fatigue. So that's one problem with these vital early warning systems that are completely-.

Gaurav Parchani:

So a lot of false alarms also or these vital early warning systems that are completely a lot of false alarms also, or at least in some cases. So you will definitely catch a patient who requires care, but you will also get alerts on patients where you they do not require care. And that is also primarily because every patient's body is different. Everybody's baseline is different and if associating one early warning score to uh all everybody, in terms of just vitals right, is going to give you, is going to give you this yield only, so you are saying that the alerts should also be sort of personalized as much as possible.

Gaurav Parchani:

So that's the next step from this, but what is the step one step beyond that as well? Right, you must so. Heart rate you would get one every minute. Respiration you would get one every minute. Blood pressure one, you would get every few minutes. Sputum you would get one every minute. Right. Now why are these four values? Only four minutes, every four minutes? Right, only four values.

Gaurav Parchani:

Because what you're essentially doing is you're averaging all of this out and you're representing this is how the body is performing every minute. Right, and you're losing a lot of information in that. And this is where the next generation of AI comes in, where I don't want to lose all of that information. For example, my respiratory rate being 35 does not tell me at all whether it's shallow or deep, whether I'm breathless, not breathless. Have I increased my effort of breathing or not? Right, but a sensor under my mattress capturing all the vibration can actually capture that dimension as well. So I'm getting the entire signal, I'm counting the number of respiratory cycles, reporting it to a doctor and throwing the other information out, because the doctor cannot understand all of that, or the healthcare professional cannot understand, can understand, but cannot consume all of that information. Right? Imagine for thousands of patients and hundreds of thousands of patients. Right, it's very difficult to consume that information, but machine has no problem in consuming that information.

Gaurav Parchani:

So now, with all of these clinical studies, all of the feedback that we get from hospitals, we have an amazing database of patient journeys which patient came at what position, what was their comorbidity and at what point of time they crashed or at what point of time, opposite, they recovered well and went home as well.

Gaurav Parchani:

That's equally important as well. And now we can train machines to learn patterns in these hidden dimensions at the sample rate of 500 samples, a second, 1000 samples, a second right, where you have enough information there in the data to actually differentiate between this patient is different than this and this patient requires urgent care right now, and it can actually stratify there. And that is where I feel no human can actually do that part, because it's too much information and too much mathematics to do where you have very less time and this is where. But I don't feel this is going to replace humans at all. Right, it's not going to replace nurses at all. I don't feel this is going to replace humans at all. Right, it's not going to replace nurses at all. It's going to generate an alert which has to be verified, which has to be understood and which has to be acted upon by a healthcare professional itself. And this is where I feel they can come together, work together, where we can take the mundane part of calculations from the human.

Sanjay Swamy:

That, okay, I am good at the machine is good at calculations, all of that let the machine do all of that, but physically checking the patient right. What's wrong with them, are they responsive enough or not in any case there's such a huge shortage, yeah, and the best this can do is sort of approach the desired ratios, so to speak. Right, I think you know exceeding it is a long ways off.

Gaurav Parchani:

Yeah, and the next to next generation. That next generation is definitely personalization, as you said, but the generation beyond, that is what I spoke about. And, uh, again, because it's our passion, we've already started working on it and we've seen some phenomenal results. Already. I feel we're less than a year away uh, I think I would say at least six months away from actually piloting it in production settings, in hospitals. We're giving this additionally to vital alerts. See, the vital alerts are very sensitive and they're already there. Imagine in partnership. There are some other alerts which are very, very specific. If that alert has come, definitely something is going to happen. Imagine them working together. I'm covering sensitivity with one set of alert, but I'm also covering specificity and precision with respect to the others as well. Right, uh, and this is where I think, within six months or so, we'll be able to pilot it for sure this is like breakthrough stuff, right.

Sanjay Swamy:

This is not like something is being done in other parts of the world.

Sanjay Swamy:

Uh, you guys now recently got fda approved and are in the piloting Dozee in the us as well, um, but coming back a little bit to just as a matter of, uh, the industry's readiness to accept some of this stuff, right, because it's also ultimately, you know, you're dealing with people's lives in the more literal way than, say, in financial services and fintech industries, where, okay, somebody didn't get a loan, you know it may have a financial impact on them and their income generation might be curtailed.

Sanjay Swamy:

But here you're actually talking industries where, okay, somebody didn't get a loan, you know it may have a financial impact on them and their income generation might be curtailed, but here you're actually talking about their lives itself, right, literally. So, you know, and and plus, this is not an industry that has been great at adopting tech in its core right. It is adopted tech in its operations, but not really in the product, in the end service itself. So what has the and not to mention this entire, you know, fear of AI and things like that, which is, you know, core to what you do at Dozee, right? So how has the industry been, you know, open, or willing to try some of this out, or to help, you know, co-create it in some ways? And are they seeing this with, you know, with slanted eyes, with suspicious eyes, or are they saying, wow, this could actually really work.

Gaurav Parchani:

So uh, I can, I can, I can say. I can say that with a little bit of personal experience as well, because I have a doctor at home, my wife's a doctor. Uh, doctors in general are more skeptical than normal, like than other professions. Uh uh, they also have a very less amount of time, uh, to actually engage in any sort of uh uh conversation. That may require a little bit of depth, right uh, and that is not from their own field, for example right, they.

Gaurav Parchani:

They have a lot of medical education going on, even in, even like. I've seen doctors with 20 years of experiences attending CMEs and learning something new, so that learning component is always there, but something that is alien to them. Technology, for example, right. Or AI, for example, right. They approach it with a lot of skepticism. It's been hard working in such an environment where you get such a less amount of time, uh, and you have to convince somebody that at least give it a shot, right? Uh, they're not going to adopt any new thing in one shot, and this is where what you have to figure out is you have to give them experience, demos enough time, show them data on their own patients, right? Or in their own practice, uh, and that is when they start looking at okay, this seems interesting. Let me take more interest in it. However, I think the holy grail of adoption in healthcare and tech adoption, or any sort of adoption, is real world evidence. Peer-to-peer learning is amazing in this particular industry. As I said, doctors with 20-30 years of experience are still sitting in continuous medical education cme events is as they call it and they're learning new stuff, right, whether it be new implants or be new diagnostics or whatever, right? Or even with Dozee, we do a lot of comes for them now and, like any other industry, continuous medical education so it's essentially their acronym for ongoing classes or ongoing education events, where they get to learn new stuff and which is usually taken by another fellow healthcare professional. So usually a doctor who's experienced a particular solution or a tech or something else, has some experience with this, has enough confidence that understands it and can speak about it. There are many cmes that happens in workshops that happen. Sometimes it's within the hospital, where they do it every week or every month, and sometimes it's intra-hospital, and a lot of events and workshops that essentially happen. So this is a very good forum and, like every other industry, you have early adopters here. Now these early adopters are the ones who are very interested in technology. Now they also approach technology with skepticism, and with AI my experience has been they're super interested With all the buzz around, whether it's from a point of fear or whether it's a point of expecting that it will do everything Like I've seen the entire spectrum right.

Gaurav Parchani:

One was we were working on a research. Obviously, I wouldn't take names, names, but I was working on a research project and I got told by a very senior doctor that why do you need to do feature engineering, right? Why just give it to the model. It will figure out on its own. And that case I wanted to say that then there's no need for me here. You have the data, you have the patient, there's the model, just give it the data, it'll work on its own right. But that's that's the level of and they're amazing senior doctors, right? I would much happily give trust my life with them, right, and again, in the same sentence, I'm essentially talking about their understanding of technology. To be that shallow, however, when we checked with them and a lot of doctors that would you be interested in doing a little bit of deep dive, not from the perspective of that you start coding from next day onwards, but from the perspective of one. You start understanding it right and second, you start understanding how to evaluate it right. There are so many journal papers. There are so many papers that are coming out on AI, coming out on AI in healthcare.

Gaurav Parchani:

Not everybody is following the best practices. Right, as in, you keep a separate testing set, then a separate validation set, right. Your testing set should never see your model, for example, right. Or a data center should also never see it right. Uh, basics related to it, basics related to bias, right as the data set well rounded enough or not, right? What sort of precautions have been taken for that right? Is this tested thoroughly or not, right? Uh, it's very easy to get to 99 accuracy when you're not following the best practices and publish a paper.

Gaurav Parchani:

And this is where doctors are very interested to understand to okay, what is good, what is actually not good, right? So when we check with them, they were super excited about it. Yes, we would love to do that. So we actually developed a course for them, an extended cme, so almost like a six-hour course including a. So the course essentially covers the end-to-end development cycle of ai in healthcare and we've taken an example from their own field which is nothing to do with dosing. So we've taken a single edcg electrocardiogram and then we've shown how we can detect afib, what apple does, did from their apple watch, right, and we showed how we can get to six, ninety six percent accuracy in just six hours as an.

Gaurav Parchani:

Obviously we've rehearsed everything. We have the model ready and everything, but I'm sure, like it took us like, I think, not more than 20 hours for our engineers to actually develop the entire course and everything, um the material to the graphs, to everything. So we cover that end-to-end life cycle with that example and with each point in the life cycle we show them. See, this is how you clean data, this is how you remove biases, this is how you select models, this is how you evaluate. Actually, this is where the model is doing good. This is where the model is doing bad, where it could go wrong, what to do when it could go wrong? Right?

Gaurav Parchani:

so you are running this course now as a program for doctors and yes, so us, along with Dozee, along with iit indore, which is my alma mater, proudly speaking about that, uh, we've combined together, coincidence, yes, so, uh, a couple of professors from there, uh, and our, our indian engineers.

Sanjay Swamy:

We've collated the course together, uh, and we're going to do the first course the doctors who did not pass je and ended up becoming doctors are not going to get an education from it anyways, so I last to.

Gaurav Parchani:

Last week I met a very, very senior oncologist, uh, who pitched his idea to me like a new idea, and it was so amazing to have a reverse pitch that I have this idea how, what would it take to build this idea right? Or how, what, like, what kind of technology would it take? Or something like that. And it was amazing to engage on that, uh. And he mentioned that, uh, I am an, I was wanting to be an engineer, accidentally became a doctor like 30 years ago or something like that, and he was super excited about the course. And imagine, like at the age of 60, uh, being excited about something new that you want to learn from scratch, being such a well-respected surgeon, so what you?

Sanjay Swamy:

are trying to do also is sort of demystify it to them, right? I mean say, look, this is actually science, this is not magic, it's not artificial, it's real.

Gaurav Parchani:

So our hidden agenda there, in which we wrote in one line and it's not really hidden, we actually talk to them about it is that skepticism is good, keep it. But you want to turn skepticism into curiosity. When you are skeptic about it, you essentially reject it at face value, but when you're curious about it, you ask the right questions. Right and never. Obviously you should never accept something, especially in healthcare, without being sure about it. But that's the difference between skepticism and curiosity. And we want, with this course at least 80 to 90 percent of people are taking that course we would like to turn that skepticism into curiosity. With a young startup in healthcare approaching that doctor, the next time will actually get more bandwidth and more uh, uh as an interest uh from them when they are actually pitching their product that's very cool.

Sanjay Swamy:

Very, very, very, very cool it both is. You know, I think it's important for success, but also it's very important for the industry, right? Because this is the future and and there's just more and more going to be coming at them. And if you don't understand the basics of it, then you will.

Gaurav Parchani:

You will just approach everything with suspicion If you look at what's required to build a good model right, or what's required to build a good AI model. Yes, there is the engineering component of it modeling, hyperparameter tuning and all of those things right but a lot, a lot, lot of it depends on the kind of data that you're capturing right, and the kind of variability in that data, the kind of diversity in that data. We as a country are at an amazing place where we have 1.4 billion people. We have 2 million hospital beds. I don't know how many million opd patients visit every day to different hospitals. I can name a few hospitals where they have 12 000 footfall per day, right, uh, my wife works at a hospital where fellows from uh belgium, from italy, from uh lithuania are coming and visiting and in three weeks they are looking at surgeries that they look at one year there the number of surgeries. So we collect and we do healthcare at crazy scale. If only we had the way to actually format the data or actually streamline data collection in a way that we digitize it, we tag it properly. It could power so much of next generation of health ai models actually coming from india not just monitoring, not just early warning system, but imagine imaging.

Gaurav Parchani:

There's no reason why, uh, indian startups or indian companies per se cannot build. Uh, because we have the manpower, we have the engineers, we have the doctors were very, very amazing. Indian doctors are very famous across the world, by the way. Uh, we also have the manpower, we have the engineers, we have the doctors, who are very, very amazing. Indian doctors are very famous across the world, by the way. We also have the large, diverse patient population as well. What we don't have is the framework and the structure to actually have this data together, and we also don't have the skepticism around, like you can always de-identify patient data and actually contribute it for, uh, generation of these, a lot of ip, and that ip will come back and help us itself, right, uh, and this is where we lack. So, if you compare us to some something like a mayo clinic or cleveland clinic or emory university, uh, they have large, large databases of millions and millions of patients that have been with them. Now, right, so somebody who partners with a Mayo Clinic?

Gaurav Parchani:

now yes, somebody who, yeah, so, from birth to death. Every test that you've done, everything that has been done to you, every surgery that has been done, every report that has been an imaging report or a blood test report or something like that, is actually a part of that, de-identified, completely Personal information has been removed from it and it's now available for people to build. So imagine somebody partnering with a Mayo Clinic starts at a much higher advantage. We have all the ingredients, but we don't have the advantage.

Gaurav Parchani:

And this is where my vision is that at some point of time, somebody should and with the health stack and everything coming along right, I really hope that large scale data models are available for people from india particularly, and preference is given to indian companies to actually uh, give it one shot to building solutions for india. And when we actually build for india, we've shown that we build for the world, right, even right. We've launched uh beyond india. We are present in uae, we are present in africa and, more importantly, we're presenting present in the United States as well, right. So it's not only for India that we've developed this right, it's for patients around the world, whoever needs care. We are there for it. Wonderful.

Sanjay Swamy:

No, I think the key point you're making is just the combination of the willingness to adopt new technology, the scale at which we need to solve these problems and the cost structure at which we need to solve these problems. And the cost structure at which we need to solve these problems are all coming together and that can serve high value, low volume. That can serve low value, high volume in terms of monetization capabilities, but solutions coming from a very low volume, high cost sort of framework are very hard to adapt, whereas the other way around is possible. And we're seeing this in other areas as well, right Right from Aadhaar and UPI and all of these as well, absolutely Great Look. We can go on and on.

Sanjay Swamy:

One quick thing I wanted to touch upon, at least for viewers and I'd like you to maybe give a 30-second view of it is you have to do fundamental development of an idea to a technology prototype, show it to people in the healthcare industry who are not really, you know, used to anything other than certified products, find people to part of it to do clinical trials here and then actually publish papers on it. You know, get the product certified, then get into, you know, large-scale deployment, which might, in our case include some, you know, manufacturing and then continuous improvement, right? So this whole cycle. Like anytime, you have a new feature, one of which, for example, is this non-contact blood pressure, right? Maybe we should also talk a little bit about what that is, because that's also a breakthrough here. You know, how long does that entire cycle take and how have you sort of navigated this thing of getting some early adopters to you know, help you with you know, trialing all of these things?

Gaurav Parchani:

um, yeah, so I have. In life sciences, uh, there is this full timeline of trl1 to trl9. Trl1 is when you start with the idea, then you have prototype, then you build uh around it, then you test it. So these trls and technology readiness- stands for technology readiness levels okay and these.

Gaurav Parchani:

These have explicit definitions of where, what you do, what. Uh, we're not this, we're not bound by regulations to follow any sort of naming convention there. But I think largely what you said. These are the phases that somebody goes to on an average. In my experience it generally takes three years from an idea to actually getting to final large-scale deployment including incremental new ideas.

Gaurav Parchani:

Yes, oh, incremental new ideas could take. It depends on the idea. So we'll talk about non-contact blood pressure, uh. But it depends on if the idea requires hardware. It's significantly longer time, right. If it's just software, it's slightly more easier. If it's AI, then it can also be longer, because regulations are still developing and catching up for AI, right. In fact, if I'm not wrong, between 2011 to 2023, the FDA did not give more than 40 approvals for AI 40 in those many years, right for ai 14, those many years, right, uh.

Gaurav Parchani:

So, uh, in all in all, it's it's a hard and it's a long patient cycle that somebody needs to be very patient about, both the founder as well as the, as well as the uh, the partners and the investors that are participating in the company. They need to be well aware that it's going to take time. Um, so, yeah, end to end on a roughly on an average to large scale deployments, you could always do, uh, alphas, betas, paid pilots, paid commercial pilots and all of those things before that, depending on your readiness, but roughly I've seen close to three years is a good time. If I knew what all I knew today, I'm sure we could have shaved off at least a couple of years from 's timeline as well, uh, I'm sure with mudat as well. Right, both of us what we know today, uh, and that's why you would see the incremental people who start that sentence.

Sanjay Swamy:

If I knew then, what I knew now, what I know now, ended with saying I would never have ventured into this.

Gaurav Parchani:

I'm glad you're saying this I think at any time I wouldn't change anything at all in terms of doing it.

Sanjay Swamy:

I think we will just be able to do it faster, super, so so let's spend a little bit of time before we close on this incredible invention of yours, which is non-contact blood pressure right, which doesn't exist, has never existed, and it's such a breakthrough.

Gaurav Parchani:

Tell us a little bit about it A little bit of like if, if, if you have a couple of minutes, a little bit of history of blood pressure, right, so, uh, blood pressure is not that new of a concept in the large history of time that essentially you think about. Right, uh, in 1700s or so, it was first time there was an english clergyman who was very curious about why is the blood flowing like? What is the pressure difference? Is there some mechanics associated to it? Right, and out of his curiosity, there was a horse dying. Uh, he wanted to like, he had some ideas around it.

Gaurav Parchani:

So in the artery next to the, the jugular artery, he put a nine foot, nine foot tube, a glass tube. Right, and the blood rose in it against the atmospheric pressure. Now, with every heartbeat it goes up and down. This is your diastolic and systolic cycles. It keeps going up and down. Right, if your blood pressure is 120 80, that essentially means that during a systolic cycle uh, when systolic cycle is when the blood is rushing out of the heart, so when it's passing through your heart, your blood pressure or pressure in your heart is 120 mm right of mercury. And when it's not passing, for example, when it's filling in the heart and it's not passing, for example when it's filling in the heart and it's not passing through the heart, then it's diastolic phase, where it's 80 mm. So for a normal person 120 by 80 is what you get the reading. So that's what the person saw in the tube.

Gaurav Parchani:

Almost a hundred years after that, for the first time, somebody actually tried it on humans, how A French physicist, if I'm not wrong they put a catheter in the arterial line. So usually when you go to hospitals and you get admitted hopefully listeners of your podcast haven't gone through that experience, but if they have, they would know that there's an IV line that is inserted so, which is basically for any sort of medication to be given to you or something an intravenous line is going through Artery is slightly more deeper than the veins, so that's where usually people don't prefer arterial lines because it may lead to infection.

Gaurav Parchani:

It's risky only in riskiest of riskiest patients it goes right. So in 1800s or so, first time a french physician put an arterial line and then put a cartograph on top of it where, with a pencil on the graph, it's basically plotting the pressure that's coming. So you get to see a pressure wave. Now you have the exact same technology digitized and you would, on a patient monitor, see the waveform when this systolic blood pressure goes up, diastolic comes down, then goes up and then goes down. And then, 100 years after that, somewhere around in early 1900s or so, a German physician finally came out with a non-invasive, a cuff-based technique when today also, you see the sphygmomanometer in hospitals where the nurse, the manual one, the nurse or the doctor is pumping it and the mercury is actually going up and down, right. And then finally, I think almost 50 to 100 years after that, we have the digital one now with every one of us at home, right, basic problems with these. These are amazing inventions to be able to measure blood pressure. In fact, you'd be surprised to know when the first time in 1800s, when the RTL blood pressure was measured coincidentally, all the patients that it was measured in were nephropatients had problems with their kidneys. So the first, for the first few decades of blood pressure, it was thought of an indicator for kidney diseases, which is true. But blood pressure has so many different other contraindications as well, or so many different comorbidities sorry, contraindication would be the wrong word so many other comorbidities as well. Right, and so the first few decades, it was considered that if your blood pressure fluctuation, your kidney is going bad whether it was or not, it's a different story altogether. Um, now, the obvious problem with invasive blood pressure, uh, is essentially around the fact that, uh, it is invasive, so you have to put an arterial line, which is something which is not advised. So only in 5% of ICU patients today it's done so. In a hospital there would be 200 patients, 10 of them would be in ICU and only two of them would have an arterial line. That's the sense of it.

Gaurav Parchani:

For everybody else, you have the cuff. Now, cuff is a good equipment if handled by a professional. It should not be loose, it should be put in a particular position and that is when you would get an accurate reading. Now, if you take the automatic once and every 30 minutes, it will take one of your readings. The curve sometimes becomes loose, right and it's more importantly, even if it was super accurate, it's super uncomfortable to live with. I've personally tried a few times sleeping with it to collect data, obviously for , and it's very difficult to sleep with it when every half an hour, something is uh, inflating on your hand and then deflating again and pressing your hand again and again, right, uh.

Gaurav Parchani:

And this is where, if you ask any doctor on the, I'm willing to bet that, uh, out of the five vitals that are there if we ask them, we can only give you one. You have to let go of other four. They will choose blood pressure always. And this is where we started researching into this. We read tons and tons of research papers and this is where I must give you credit as well, because I remember you sending one research paper as well which contributed. I couldn't understand it.

Gaurav Parchani:

Yes, so, after reading it a lot, we experimented a little bit with data and we understood that with heart rate, respiration rate and everything else, we are only dealing in time domain. We are only dealing at what repeats. If a heartbeat is repeating again and again. I just need to count it so many times in a minute and I can say I counted 70 beats. That means your heart rate is 70 beats per minute, it's 75, then it's 75. We are not worried about what the beat itself is. If it's a heartbeat and if it's repeating, we are counting that.

Gaurav Parchani:

But with blood pressure, we need to now start analyzing that beat, with what intensity the blood rushed out of your heart to the aorta, because our dosage is right under the mattress around that area, around your upper thoracic region. So it's actually carrying that impact information as well, right, but because it's passing through your body tissues, it passing through the mattress and everything, it's very difficult to get an absolute value to that. So you can easily not very easily but you you have so many metrics in the vibration data that can essentially show you how the blood pressure is changing and that would be captured in the vibrations that are generating out from your heart. And that is where we built AI models With tons and tons of data Collected from ICU patients With arterial blood pressure, because that's the most accurate for continuous measurement.

Gaurav Parchani:

We trained machines to identify changes in blood pressure and even subtle changes in blood pressure. And then we said you need one calibration value To begin with, because we cannot give you an absolute value. So when you calibrate it with respect to another machine, once for a patient, once from every 10 minutes, then onwards you will start getting the change in blood pressure and automatically you will get the absolute value because you have the beginning point of the starting point. Now, this is something which does not exist in the world. As you absolutely, very, very rightly said, there are many that are coming which are cuffless, where you have to put on a finger or it's worked through a patch or something like that, but non-contact. We are the first ones in the world to do.

Gaurav Parchani:

The regulatory submissions are going on. The clinical studies have shown phenomenal results. In fact, we've published partially the results as well and we're publishing more results as we can get these results out. But this is a game changing feature for us as well. This was something that took almost a year and a half to build, to perfect, and then, after deployment, it required at least. I think. We are at version 4.0 now and we've had three major changes. And then we had some few minor changes in the middle as well, where we've improved the accuracy. We've improved the specificity and sensitivity of the model.

Sanjay Swamy:

Perfect, and just to clarify, it's live in India, but the regulatory stuff is really more from an FDA perspective.

Gaurav Parchani:

Yes, absolutely Awesome.

Sanjay Swamy:

Great. So, look, this is a crazily fascinating I guess topic, basically fascinating I guess topic. And you know, really you guys have done a lot to be on the cutting edge I guess bleeding edge is the wrong word to use in the healthcare space. But kudos to you for staying the course and, of course, the best is yet to come. So all the best and look forward to the journey ahead.

Gaurav Parchani:

Thank you. Thank you so much for having me. My pleasure Dear having me Our pleasure.

Prime Venture Partners:

Dear listeners, thank you for listening to this episode of the podcast. Subscribe now on your favorite podcast app for free and you'll be the first one to know when new episodes are available. Just search for Prime Venture Partners Podcast in Apple Podcast, spotify, castbox or however you get your podcasts Then hit subscribe and if you have enjoyed the show Spotify, castbox or however you get your podcasts Then hit subscribe. And if you have enjoyed the show, we would be really grateful if you leave us a review on Apple Podcast. To read the full transcript, find the link in the show notes.