The Hardtech Podcast - Ovul Episode
Guest: Serhii Zatsarynin, Founder & CEO, Ovul
Hosts: DeAndre Harakas and Grant Chapman
Recorded in person

Note: this was recorded in person and the source only diarized two channels,
so speaker labels are approximate and blend the hosts and guest in places.
Names in the raw audio were also auto-transcribed and are unreliable.

========================================================================


[00:00:00] Speaker 2:
Welcome to the podcast.

[00:00:08] Speaker 1:
everybody, welcome back to the Hard Tech Podcast. I'm your host DJ Jericho's with my co-host Grant Chapman. Not usual suspect. Not. Not the usual. Look, I've gotten so much grief and feedback for. He's been the usual suspect for all of, like, 100 episodes. And so at this point, I think he's more than just he's he's officially the co-host. Yeah. So you've been promoted, Grant. I hope you're excited about that. Yeah. Fortunately or unfortunately. But  no, everybody's super excited for this conversation today. We actually get the privilege of doing this in person. So thank you so much for making the trip, bringing your family and hopefully make maybe get some good steaks after this as well. Absolutely.  Well, again, welcoming Sherry, the CEO and founder of Oval to the podcast. Welcome.  Pleasure to be here. Thank you so much for having us today. Absolutely. So we have some physical product in person. Can't wait to have a chance to show that off a little bit. But we would. Yeah. Would  just tell me what's better. Oh yeah. Totally. I would love to kind of kick it off on just a little bit about you and your background. I think you have such a unique story just on how you got to starting Oval. Yes. So my whole my wife, I was in diagnostic and funny enough, my grandmother and my mother are doctors, and I told them that I will never be in medical space. Right? Never, ever.  But since the fourth grade of my university, I joined  Siemens Health Seniors. The way that company Simmons is one of the top diagnostic company. And I ended up running the whole business in Ukraine for in-vitro diagnostic for the last  five years. Before all of you and  why I'm here and why women's health  tech and testing the and the hormones at home. So imagine this in my work environment, I built the top ever factories of top notch diagnostics at home when we were trying for our baby. My wife gives me her cup with her urine and I put like, strips inside, like testing for chemicals, but it's just been in a cup. And she was my I like  if for like when I moved in this business space like every new human which is in this involved in the production of diagnostic is like a mistake. You know, this is not a potential mistake. Yeah. It's like it's like you need to avoid this. But at home I have only options. You have to watch with my eyes. And they say, no, like, this is like, why? How this might be possible. Like, let's change it. And then the subject of you love it, love it. And so, you know, what would that transition point for you of like, let's start this thing and let's like start. How do you start testing? Like what was that transition to like when the idea was, yeah, I want to work on this. Like what was the first thing that you tried? Yeah. So  okay. So  at home we were testing all this stuff and trying for a baby. And the colleague of mine, who is the embedded engineer at Siemens at that moment was sitting in his desk, and he was using a lot of different stuff and creating the PCB or something like that. And I was like, wow, it's kind of interesting. I can ask this guy, but what he thinks about, is it possible to create a small microscope for at home use and how much will it cost? That was my question. I thought I'd come to him and say hi. So how do you see, like how much would be my perception was that it's like around 300 bucks or something like that. Very expensive, very expensive. It's like a big tech AI microscope. And he tells me, like I would say like less than 50 bucks. Yeah. I say, okay, okay, I have 50 bucks. What? Where do I put this? In the venue. How are we going? I get it, yeah, that was the starting point. That's awesome. That's incredible. You know, one of the first things I found so fascinating about your story on Building Oval  was really around. We had our first intro conversation. I was like, oh, you know, hardware is hard. It's usually part of the conversation.  And you actually came back and said, hardware is not hard, which I thought was incredible statement. Could you kind of talk a little bit about that? Yes, I like I was driving here for like ten hours and I was thinking about this specific line like, why? Why a game was this and how this looks like. So I can say that  before it was hard right now to create a prototype. It takes quite days or weeks. So this is a time when you can deliver something which will be somehow working. And that's what we did it. Although it took us 30 days to create the first prototype, it was really working and we started delivering to the users like in in about two months. When we decided to start the company, we already shipped the first user like like a pilot units to go get test feedback on  the first one and they go like, very bad. They were, you know,  they did exactly what they needed to do, which was see if the market liked the thing and they did nothing else. Yeah. Or it worked for a little bit. It needed love, you know, reset. Plug it back in, I would assume. So we. Yeah, we ended up having 52 product versions, two CDL production. That was our very own pathway.  But hardware is hard and not hard anymore. So what was admiring for me during the last years the Consumer Electronics Show last year, I would say that about a third of all best innovations go 3D printed, right? So meaning that it's not about the perfection or it's not about something like making a very great product that looks great and amazing. No, it's like more about the speed of production. How fast are you going? And with that, with this prototype, you are able to get the best ever innovation. It says, I think this is how software did it. So I think this is like the like the Renaissance software went through when like the first, you know, there's the first internet revolution in the late 90s, but then there's a really fast like app revolution of like the, you know, post 22,006. And after that, all this stuff just really iterated. And you'd have these apps launch and they can make them so fast, they would try wacky ideas like, who wants to rent out part of your house to a stranger? Airbnb? Who wants to drive strangers round in their car, Uber like, you know, these things that were wacky ideas were fast enough to try that software allowed people to make these weird software products that ended up becoming part of the things we love today. Hardware started this journey way back when 3D printing kind of broke through with like, you know, the first MakerBot when they let you know open source things like that. And then Arduino and Raspberry Pi development kits for electronics started making electronics easier, and 3D printing kept getting cheaper. And I think that that revolution allowed people to go from 0 to 1, like you just said, like. And then you can try things and get market feedback because it isn't actually making it perfect is necessary to get market feedback, but getting really good market feedback is necessary to make it perfect. Yes, I think that's an excellent point, but I think to kind of go back to on your side as well, kind of talk a little bit about how, you know, there's a lot of founders and product owners and leaders who say it needs to be perfect before I launch it. Right. I need to make sure all those iterations you have before you put in front of a customer, and then they're going to love it. You went the complete opposite approach. And so maybe talking about the outcome of that would be interesting. So you know you have path A, which is, hey, we're gonna get this prototype out in two weeks and have 30 iterations versus, hey, we're going to go all the way to the finish line and then get in front of a customer on the back side of that. How did the customers respond after iteration 30? What did you learn along that journey and how has that impacted the growth of the company? Yes,  I still think this this was the right approach. Like  I  explain why and what we have right now and why it happened.  So we had a very bad experience with our customers always, like I'd say. So we have an egg shaped device where you need to like,  like open it on two on two parts. So they were broken parts instantly. Always. Whatever we do, we put like a sphere or you just go down or up. It's always the damage in the parts. Like, you know, the first five iterations, like, like that. It was physical damage, like in order to open it and put saliva inside. So later on, we understand that how they wipe to have it, how it's,  you know, where they store it, where it's like on a nightstand or it's on a table how they use it in their life. So we changed a bit of real design.  So we want to wanted to avoid the buttons. Like we don't have any buttons on our devices, something we don't want to use. Customers want to use buttons. So they want to see the screen. They want to see the connection. They want to see how this is working or not. So we we ended up putting some light indications in different scenarios. So we tested it out.  Cowering scheme. We were asking customers what colors they prefer, what the touch feeling for the sticks they prefer. What color material finish right. How does this look and feel in my hand? Yes, it should be like delivering the whole experience and a part of it is for sure how you view it. Like what do you have in your hand when you want to have a baby? You don't want to have any distractions or more distractions because you already  have stress. So we have  asking that and the changing. Like every time we were not listening only and say, okay, so wait around, we were at it or something. No, we were just delivering. So how about this one? How about this one? And this is how we ended up building it like very fast about like one iteration in two weeks. That's that was what was happening at that moment. That's awesome. And so you know you're iterating this product. So for all the listeners I have the benefit of having seen the product talk to you before. Can we jump into what the product is  and how it works? Because I think that, you know, everyone thinks you must be building something very simple, right? That that has no complexity to it. I know what you build, and it has a lot of pieces that have to work together to make the product work. So, you know, we're we're trying to fix fertility testing. Right? So no more sticks in a cup of urine. What have you guys been working on? Yes, I have habitus  on the desk days. It's like a different generations. And, this guy is something which was already on the market for many, many years, and it's like saliva microscope. Okay? It's a real microscope where you have some kind of slides or stuff like that, very big saliva. And then you need, like, somehow put it on, on the light and then teach yourself to track specific patterns, which corresponds to estrogen hormone changes. So you get a little guidebook that is like if you don't see the guidebook, it's like. Oh, if I don't see this shape, it's not it, okay. It's kind of like reading tea leaves or like the stars. Yes yes yes yes yes. And one of the main constraints why this technology is not so  common and not everybody is using it. So you need to teach yourself like invest in a one week to week at minimum to understand those patterns. Can you see this or not? Then the shapes the leaves.  So my idea was why not use image recognition? Basically AI pattern recognition and AI. Good. That's it. Like make it simple. It's and  the generic of this product is it's totally reusable. So I like about this you just whenever you see you just do this. But we took  away this human factor when you need to see it and to wait and then recognize the patterns. So we just simplified it. Yeah. What you like, if you get the human out of the process, it works a lot more often. It's weird how how not good at following directions. Humans are, okay. So my my seaman's wife,  but  I was starting this  as an engineer, so I was installing the physical people. Then I was the,  the field application service. When you were explaining how the test is really working, and I have a stats, like 62% of all errors are human related areas, but this is like a top notch diagnostic. Everybody is trained well enough. They have higher education. But no, it's like no, it's still here. Yeah. That's awesome. And so when it comes to like the technology right. So you said with the saliva and I mean this technology is not new. You guys just applied it in a very new and novel way. So I'm curious there, what is the true technology to be able to measure the saliva and see what's in it. Yeah. So it calls saliva. So this car basically looks like the very last device and the whole solution. So now it's still a microscope, but it's a digital microscope where you open it, you put saliva on the slide on the bottom and then you close it. That's the whole test. That's magnetic, isn't it? But basically you wipe out the previous saliva with any piece of cloth and then you put a new one. You're by the new one. That's the test. So you have users could test work several times a day. Right now,  the technology itself that calls saliva, whenever proteins and set in solution are combined, they create very specific structure which looks like a firm flower, like a  firm leaves. Yeah, it's like a snowflake. The builds off itself and has little branches that come out the sides. Right? Yes, yes. Crystals or something like that. And it was discovered in 1945 and a lot of,  clinical trials and that was already there. But it was still very hard for the user because you had to be an expert to read the literally the pattern in the tea leaves, but in the saliva crystal leaves. Yes, yes, yes, yes. And also like the way how you use it,  I think that was a part of our iterations. So we understood what's the best for the user, like how they have their daily routine, how they can implement. Where do you fit in? Yes. How can we fit it in that? Not to spoil away. So just make them do something extra then they could do with our device. We have the same theory on like safety devices. When you're making a safety device, it only works in an accident. So most of the time for the user they see no benefit. So you have to make the user experience be so seamless with their day to day that it always gets used, because the safety device that no one uses doesn't save a life. But same thing with your test. You have to make this test get used because they don't use it consistently. It doesn't provide the data they need. Yes, I think this is one of the most like greatest stuff that we accomplished, that we made it so simple, like five 10s when you whenever you see the device, you're able to test and then they start testing and they just test, test because it's easy. But if you need to change the whole day, the whole routine and come back from work and then do some extra instead of an alarm, because if you don't do it this time, the data doesn't work. Yes, right. And so on your device. I spit on my slide, let it sit down and I can walk away. Does that store a certain number of samples of data and then I think it with my phone. How does the data leave the device? Yes. That's the reason why we use the Wi-Fi to transmit the data. Yeah, yeah. So you need to wait for about 42 minutes. And that's why I'm so interested in guys in your agreement and the Chamber of Utilities. So  the idea is that it with different humidity takes different time to have the sample sample. Yeah. So in general we have devices all over the world. And we understood that it's like 42 minutes and it's enough to drive.  So you don't need to wait near the device. You just put saliva and then you go to your business. And then you have an app and  result is transmitted through the Wi-Fi. Oh, it's like you like pair that to your home Wi-Fi. Yes. And so that's always online and whenever it's ready to uploads. Yes. Amazing. Again we saw that Bluetooth will be the option for us. But they would leave it at home and there's no more connection to Bluetooth right. Yeah. So  we want to like five ten minutes. It's like no they don't want do this. And that's brilliant. Oh I, I like that I was convinced that was a Bluetooth paired device from the entire use case, but the Wi-Fi sync makes so much more sense now. It's not easy to do. It was expensive and hardware compared to just going Bluetooth inexpensive and firmware development, but it made the user experience good. Oh yeah. Battery life.  It's like all of this stuff. It's hard for everything, but it made the user experience because it was worth investing in. That's one of the things, I think that you guys have done a really exceptional job on. I think one of the things that you quoted when we first met was just an obsession with the customer. I mean, you're back your background at Siemens and even starting the company in the first place, you had the relationships with doctors, you had relationships with people that kind of help you get off the ground there. So like when you were thinking about starting the company and as you were continuing to grow, how much of those previous relationships like fueled what you guys were able to do as quick as you did a lot? Fair enough. 99%, I think everything was because of that. So  I, iterations take track always takes time, not only on the back end or development or prototype for sure this year, for each and every thing of it.  But for kind of like human relations, customer relations, we got able to deliver fast and get the feedback because we were using the doctor support. So we use the specialist, we OBGYN clinicians, they were controlling the patients like direct patients for pilots as well, but mostly like they use their like network or their understanding how this can be beneficial for the clinicians. So they were our delivery persons and they were gathering all this first experience. That's awesome. That is so cool.  So in your in your system, when you're building these prototypes up and getting this production, you know, where was the hard part of the transition, right. Like, you know, there's the early prototypes that are probably like a few development kits in a shoebox. And then there's the first 3D prints and custom circuit boards, and then you have to start getting more and more real. Which was your, like most difficult journey between, you know, the prototype with the shoe box full of dev kits and production? Yeah. So like two parts. I would say the first prototype in yet was like my my son's microscope and the camera system on top of it and, and with  with the  no clue or something. It was just pulled together. It was kind of easy.  Then when we were pivoting, when we had like first 10 or 15 users,  then we want to have some consistency and you wanted 50 units. That's not easy to make. Yeah. And and you want to like constant the things that the same camera system, at least the same weight in the same support, same humidity levels, like all this stuff, like it took a lot of  time on our hands. So if we were able to  outsource that at that moment, I would do this. But the very frightened step where we were not like understanding what we would get is when we start the first zero production. Yes. Yeah. It's like prototyping. I, I especially I, I took this one. This is one of the very last prototypes  and this one is studio production. And I, I can say that if I can open it. So there is a huge difference here. Yeah. And this is the last kind of prototype versus this is the production version. Yeah. So not only wow I can tell you you couldn't mold that. Yeah I thought that is something like a magic as well. So I'm not I was not too much technical person like so deeply. Now you are. And since for the for the listeners that are just listening on, you know, the streaming platform, what made you say that about those? Until you start designing plastic parts, you think you can make anything out of plastic, any shape, you can make it a plastic. And that we've determined that was a lie. You have to be able to pull the mold apart. And if, let's say you had a part like an egg, where it's fatter in the middle and it's skinnier on either end, you can't pull the tool out. That made it fat out. So I instantly knew that was 3D printing because it's very easy to 3D print, which is why the production one is bigger at the bottom than it is at the top, so you can pull the tool out of it. So everyone listening. You can always look for the parts that are molded knowing that whereas unlike your water bottles, these are blow molded. So you can see the little nick at the bottom for the cameras right there. This is where they hold on to the plastic, like the what they call it with the puck. And they put a nozzle in the front and they pump high pressure, hot air into this hot piece of plastic inside of a mold that expands the plastic. And they open the mold and the bottle drops out. So that's what you can make hollow parts. Yes. That was crucial step for us. So we had, I would say, pretty decent level of prototypes and they were already on the field. And then we understood that it's impossible to make okay so what's next.  So we were searching for the expertise the like a who can help us to deliver the serial production more or less the first ones I know like from our prototype and like none of us had had this experience and even if we now have it, so we had  after this step, we had three times changing the plastics and three times changing the molds, and five times changing the PCB and stuff inside. So we got to an orbit. You have an iteration inside of the serial production where you change no functionality. It's exactly what you wanted. It just took you 11 iterations to make it at scale. Yeah. And then you have to go through  making like a test fixture for your circuit board to program and a test them off of the line like everyone forgets to build that. And that's a whole product of in and of itself. Yeah. And the technical cars. So how you can manage the time span for creating the device not be put on its side every single test, you know, it's like another story. Yeah. How do you make sure that none of them get put in a box that don't work right? How do you how do you catch that? Still figuring it out. And we're and we're back up to present moment, right.  And so one of the kind of underlying questions of this is you guys bootstrapped to the first, you know, 30 iterations on the prototype. At what point were you guys completely bootstrapped on this journey? Obviously, hardware is hard. It is expensive. And so what was your journey on the fundraising side? And like kind of what did that look like for you guys? That's kind of the best question. So you was right that now it's very fast building, but not every VC understands that. But we were probably one of the first ones who delivered the serial production product on our own. And we were raising when we already have like some kind of hardware and in market, not even prototype. You had hardware like off a line. Yeah.  Like in market. And now retrospectively, I understand that one and a half years ago when we were trying to raise for this, they were just like frightened. But I never seen this like something is wrong, but you skipped something important. I just don't know what yet. Yes, yes. Like why? This is the first time ever I hear this story. How can you be in the market? Okay, so you'll just be there, like, you know something? It was like change of the mindset, which is now happening for the hardware. It is. And  I think the other side of that, that I can tell from talking to you, our first meeting in this, the reason your team was able to do it is your background of problem solving. It sounds like Siemens might have had a hand in this, but I also know that your experience prior to Stevens was pretty cool, but your engineering history encouraged you to find the friction and remove it, to find the errors and find a way to engineer around the errors. And from everything you've told me through your prototyping path, you started out with his microscope, hot glued to a, you know, to a camera. You just took away the friction every step. And a lot of first time founders don't have that experience of how do you solve a problem one bite at a time and they get scared so they don't know how to move forward, where you and your co-founders just tackled the next problem and kept 3D printing and then found a way to get through production. Yes. So this fast speed iteration allowed us to train the eyes. So one of the companies here is for sure this the AI, which should be we have now 99.5% accuracy in many like divisions of of what we can track. And this is huge because we have the data. So we're not waiting for something or searching around without producing the data. So important we I see this today. So in medical device especially but in a lot of consumer electronics or medical device IoT internet of things is just about collecting data these days. And I see these users and these  are these founders that are so scared to make users use something that's not perfect. So they're not doing large pilots, very tiny pilots. And  on guys, how are you going to get the data you need to prove that the thing works at a larger scale is accurate. If you want to build your models, build your data set. You need this proof. You need to run a pilot in a big one. Like, how do we build a thing at hundreds of volumes, you know, 50 or 100 units, but not yet, because that's expensive. We don't do that when it's right. How many prototypes did you have in the field before you cut tooling? I would say like hundreds of hundreds. But interesting stuff here. Yes you're right. So whenever you start to produce your own data, so you see the proofs, but you also see the new patterns. Like you never can get it before you in the hands of your real users, or when like when you are in the market and you have different scenarios of user behavior. So this is how we today we have about 360 devices from the serial production batch in the field. And we started to see the new patterns. Not only the Astros generated stuff, which is like well known, but for other like scenarios,  other body responses, other hormone levels, like I can't say like exactly what we can see right now. So once we're off camera, I'll ask what. Well that's awesome. And it's I think this is so interesting because I think the the products like Aura Ring or the products like loop have normalized collecting data on ourselves. That's right. Shown the power in that. I mean, you and I talk about this all the time between the medical devices, like continuous glucose monitors that are literally saving lives every single minute of every single day by collecting data constantly to hoops and horror rings, to helping people form at their highest performance, we are getting more comfortable with data and what we are now realizing. The more data we have, the more we can see. And like we're just starting to open the door. Yeah, I think we're just kind of entering a space where  hardware was obviously not sexy for a very long time. Right. And SaaS just, you know, sucked all the air out of the room for insert a decade.  We had the mobile app revolution and the SaaS revolution and things like that. But especially now with AI Claude code, these different coding platforms, we've kind of  democratized the ability to make a software platform. I mean, I don't know how many HTML docs that I've created in the past week for different things that I do for work, but at one point I would have cost me like 30 grand, and now it's like 30 30s, right? And so from that perspective, we've seen such a resurgence in the investment on the physical products, because while all the information and data was was absorbed from the entire internet that trains these large foundation models, the net new data that's available is coming from real devices that are in the world, stuff that you're collecting right now. And so I'm curious, you mentioned like, hey, a year and a half ago, we were talking to the VCs, hey, there's risk here. But now, like, how has that shifted for you guys and what's it kind of look like  from that perspective? Yeah. So  I just changed over  the months ago. We signed distributorship agreement with  Canada. Congratulations. Congratulations later on. So this is the first time I ever talk about this,  but this is how we are  like we are going on. So we understood what kind of data we really see and stood. What is the value in this data for the clinicians, for the users, for somebody who make a decision on it? And how can we use it in real way?  And all that was impossible before. We are on the market before we are like our users appends on on that. So like everything changed. I would say like the, the way how we started, we got for my own family problem and we were having this child and then we were trying for a baby and basically to create an overview as an arbitration detection. That's why the name was from ovule. So then ways around, we understood that we can support women in perimenopause when the estrogen is to actuate in this chaotic is just not only estrogen but all hormonal levels. And we can create a tool which allows you to understand why you feel yourself like that, track real biomarker, and then maybe act on this in terms of different insights into your body, mind and nutrition. Right? So that was a change again because of the users how they are using our product. What can we do this with that weight around we understood we can support menopause transition when the periods are over. And you use in some  some kind of  hormone replacement therapy or not. So we can support those users. So now we see more and more and more. And today the way how we peach and the round how it how we do this is it's totally new startup and it's totally another startup  doing it one and a half years in the startup. Isn't that right? How the company at the onset of what you go out to start versus the company that it evolves into at scale, is almost never the exact same thing. The first company I ever started, you know, I tell people, oh, my coaches use a lot of apps from the sports team. That was the later iteration. It was actually because I really struggled to learn the playbook, and I wanted to make an app to help me, teach me how to do it. As it turns out, coaches didn't want to pay for that. They wanted to pay to help themselves out. But that's just a prime example of just continuing to listen to the customer. And as you get more iterations in front of them and get more user feedback, now you guys have unlocked a significantly larger market than you would have otherwise thought at the onset of the company. Well, and with this, you're getting the window to new data that's going to open doors to other new data.  I mean, this is the I'm going down the rabbit hole of like tracking how far you guys can go. The question I've got with this is hormones are notoriously hard to track because usually it's test strips and chemicals and difficulty. Are you guys working on trying to find other ways to correlate, like clinical data for other core moans or other things you can test, you know, using the old school methods in a small sample set to build an up image data, to see other things, to see other patterns. Is there any way to collect data from a subset of users that is already collecting that somewhere? Right? Yes, for every question. I'm a big believer in data, so the more data we have, the more correlations we can find and track. That's what I saw my wife being in diagnostic in the medical family. Like when you have the data you can act on it. So if we do a work that we understand what we see and we are trying to find and map out the patterns from our users. Yeah, like find the truth source that you can align against. Yes, yes. And how we do this, we educate ourselves and also we are hiring the board members I would say. So we have now  Mark Snyder from Stanford Web of Mike Snyder name who is the  genomics is now joined us. And he is helping us to understand what what really we see why why this is how this is connected to other areas we can explore. So very recently we understood that there is a whole bunch of, of things to look for. Yes. And it calls reproductive  restrictive reproductive medicine. Like before April this year, I was not aware about existence of such kind of a huge amount of data and trials and people who are working in this field. So we worked in that direction. And yes, now we can map out using their knowledge and, I don't know, years of 50 or 60 years of data that they have. And the data that we see from our users on their fertile health, on the hormone levels. So now we are doing this match right. And it's like you need to go find the studies that are running right now and send oval units to the studies so you can get oval images while they take chemical test data and are paying for this clinical to find out whatever they're trying to learn. And can you correlate new image data with chemical test data. So we don't need the chemical test data in the future. It's just image data. I like it as a tech arsenal or as a founder. This is like, wow, what we see is don't like it. So when I say that I have something new. Like related to this kind of,  like information, for example, or something else, they say, okay, how do I turn that into money? Yeah. Yeah. Right. What's the what's the return on that? It's been right now. Yeah. But something is so interesting to me is I literally just made a post about it today on LinkedIn. And it's this idea of, like the surface area. You and I talked a bit about it, the surface area of product marketing fit when it comes to hardware versus software. And what I mean by that is, you know, with a B2B SaaS solution made for cross, if that CRO doesn't use it with his team, there's about a 0% chance that Grant's going to pick it up for the engineering team here at Glass Border. Vice versa,  while as if you actually create a physical piece of hardware that people actually use the surface area for, where you can kind of get pulled into is so much wider, right? You and I think the caveat to that, and the add on that to make that possible, is going all the way back to the beginning, where you guys did such an incredible job of focusing on the customer and the use of and getting them to use the product and enjoy the product, the magnetized piece of it, making it simple to use, making sure that it reported easily through Wi-Fi, not Bluetooth. And so they actually got the data that whenever they did it, you couldn't get to the point where you are now today. And like taking in new data if you didn't cover the human factors part of making sure they use it in the first place. So that was the kind of the huge unlock on like,  shout out to industrial design and usability and customer feedback. But it was just an interesting the mix between those two things. Curious if you would tend to agree? Yes. Again, being here today, I say yes, but when I was building I was not aware about this. Yeah, I was I was just doing this. Who cares? I just need the data. Yes, yes, yes. So this is a good way. Like how can I get the data? How can I understand, like what we really see and the like coming back a step, step step backwards. So it's great when we have a data which no one else has. Like this is something where I wake up every day and I constantly look at all the photos of all devices just to  be aware, like what's going on, what the patterns look like. I'm basically the best experts in  dry survival and probably as of today, what can I do for us  as a company or as a person? But yes,  but this is great. So we see something that is happening and when it's happened with somewhere two, three persons, then we start to see this like, wow, this is like and that's why I love about the hardware and I love about something like getting the new types of data which which probably I believe will be incredible for humanity. Well, I think going back to the average, this is the key of good user experience design. If you didn't have that, no one would use it often enough to give you the big data to find the little patterns, right? I think that's the the magic moment. And so how are you guys thinking of,  what's the word I'm looking for? You're going to have all these images. What data are users giving you to correlate those images to outcomes. Right. I understand if someone's using it to track their ovulation, they're going to be reporting some data to you. Probably to track. Yes. I'm ovulating. What from that data set are you trying to track to find other outcomes? It's very sensitive question. And I always get this when I'm somewhere  teaching or something. So how do you use the data? And I want to say that for over we decided not to ask for a person name, email or phone number. So basically any person can use over with anonymized user. We don't track that. But what retract and what we want to have is the real data which can help other people or other communities. Research. Yeah, yeah. Like research. So that's why I ask, like  the way how we build the application application is a great site for that, which is also user experience. I'm not asking for your name. I'm not asking for your personal identifier. I just need the science so you know how to  you know what my data is telling you and you can help others. Yeah, I can say, I can't say that we understood this. Like working with users. It was our first approach. We don't want to have this. Why? So what I want to have exactly what what Asia you are. What's your weight? What's your cycle? When is what's the date of the cycle? Where are we in your cycle? Yes. When you have these kind of  hormonal levels. So that's what we were asking. And when we switch to menopause it was very new for me. So my mom was not talking to me about human,  about transition. Like it's a taboo. And it's  not even in the doctor's family. So that's why I was like searching, like was that I went on courses to understand what kind of symptoms are happening when they happen. And I was asking users, what, what do you really track? Like, what do you want to know about your own health and let you just add this in the act like a two week a symptom or  the temperature or the levels of your hormonal from the lab?  Or how much sleep did you get last night or this week, or what are you eating? Yeah. Yes, yes, I was asking for the track. And then why it's matter for you. Like why you think so? They say, okay, because I saw this and I go and learn why they saw this. What? What was the name of that? And then coming back and put it in the app. And then we have the data point, which seems to be like crucial or great for having  new errors or for you. That's amazing. And then I think once you learn that that's important, you can push that out to other users and educate them. Hey, I would you know, ovule not I ovule would like to know,  can you tell me your lab hormone levels and when that test was done. Because with that I can help you track this other thing. And it's this like give and take of yes, we were asking for more data, but look at all the other things we can do with your data, right? I think that's the the interesting balance of where we're going in like IoT and med device. I'm hoping that maybe one day someone who wants to build a SaaS tool will build a great database. That's my database for my data that I can upload data to from my band, from my  overall, if I'm doing saliva testing for tracking my hormones for any reason later in life, all these things to find the combinations, right? There are so many things I think about health that are a large stack of combinations. Not 1 or 2 orders like you need five things to happen at once to see that signal, and I can't wait to find those. Yes we have. We are working on the integration to be able to provide this data. Last couple of questions for you. One, I wanted to give you a chance to kind of talk explicitly about you said the first time you were announcing,  I think you said something about Canada. If you want to kind of capture that in a moment, happy to for the clips and things like that. If you want to talk a bit about that, go for it. Then I get your last question as well.  Yeah. So we have a first partner who is the Canadian  company working in the field of  and diagnostic, and they want to use of you as a support tool for their current patient, existing patient. Again this is a new data where which can improve the quality of life faster. Yeah. So that's what we're working on. We will be announcing it like  after this podcast the PR and everything. Then we will put everything together. So this podcast will come out exactly the day after you announce that we'll hold that. So it'll be fresh news for everyone listening.  Fantastic. And then I guess my last question, which is the same question it has to be right,  is just around. You've gone through an incredible journey, a ton of lessons learned. You come from a pedigree in the hardware space, in the healthcare space will be your advice for either a younger version of yourself when you first got started and or founders product leaders that are currently trying to bring something to market today. Paid users and pilot users are different users. I was not aware about this. Like whatever, whatever you build with the pilot users, it's not the same, which is going to happen when you sell the first device. So for that reason, sell the devices whatever you have. If you are in hardware and you think that it's not beautiful, it's like ugly, it's not open and it's broken in from time to time. So sell this one. Like don't wait like never ever. And like why? Because the user experience is totally different. And then you need to change. Even if everything what you have learned from the pilot users is not the reality.  So go and sell. Go. Even if you have nothing to sell, sell something. And when you say that, what is different about the pilot users versus the first paid users, is it like expectations that they have with the product? Or what's the difference that you think? Yes. Expectations, right?  It's going to be the same profile like it's going to be the same person, same age, same ideal, same symptoms. They're having symptoms like trying to conceive or something, but ending up with different, totally different questions. I think that's such good feedback because that's how you get. That's why people always encourage you to charge for the product. Right? Because I might talk to you and, you know, if you like my idea, you want to test that? Sure. Yeah, I kind of like it. You could do this. But if you asked me to pay $500 for that product, I'll have very different feedback for you. And I'll be very more explicit as to what it will take for me to purchase it. And what's amazing is $0 versus $0.99 is still a huge difference. There's less difference between $0.99 and 50 bucks than 0 to $0.99. Yes, one box just charged one bucks. And then it's yes, it's another index. So one of my favorite fallacies with hardware products, it's expensive right? To try and make hardware especially like 50 prototypes at a time is like the worst volume because someone at some poor guy is handling those things. It probably sounds like it might have been you,  but they're expensive to build from an expense, labor, etc. and so what a lot of founders will do is they'll go do a Kickstarter or run ads on  renders and fake products, right? And instead of selling it because they don't feel like they're ready to sell it yet, there's going to run ads and they'll do click through ads. So they're going to pay for social media ads or do traditional marketing and show their cost per click and show that it's really good, but they're not showing the waterfall all the way down to the purchase. And what I've seen some really smart founders do is build a website based on all renders and make them put in a credit card as if it was ready to order, and they'll go check to run the credit card that it was real, and then send the person to the email. Hey, thank you so much. We're actually not ready to sell this thing yet, but now they have real data. Someone was ready to pay the money versus just clicking to a  like, you know, sign up on our email list site. I think that money, whether it's a dollar or $50, it's so different to test market feedback. This is how we our next stage of development of the second product. Yeah, this is exactly the like the way how you should do this. Yeah. The marketing test. This is I think the technical term is called a painted door test. Right. If someone is willing to walk through the the storefront only to realize it's not being built yet, you can count on people to walked in. That's a good indication that the store would do well. 100% yes. If you have every single conversion rate besides the purchase data, that means you could be really, really close to absolutely nothing. You have a really great top of funnel. Yeah. And and yes, that is the lesson of the episode. I love it again. Thank you so much for being on the show, everybody. This is the Heart Tech podcast. For those of you who are just now tuning in listening to this one. We are currently in season three and really excited about it. Great. Thanks so much. Thanks for joining.