Podcast Episode 22

Navigating the MedTech Maze, Predicting the Future, and Defying Startup Odds

In this insightful and engaging podcast episode, Ryan delves into the captivating journey of Amir, the brilliant mind behind Bio-Conscious. Hailing from a lineage of healthcare professionals, Amir shares his transition into the realm of computer science, machine learning, and AI, driven by a desire to merge healthcare with technology. The narrative unfolds as Amir recounts the pivotal moment when he developed a groundbreaking prototype aimed at predicting and preventing complications in type one diabetic patients. From the challenges faced in data collection and algorithm design to the unexpected twists in partnering with major companies, Amir’s story is a rollercoaster of perseverance, innovation, and the relentless pursuit of enhancing patient care through technology. The conversation explores the highs and lows of Bio-Conscious, offering a fascinating glimpse into the world of medical software technology and the transformative potential of predictive algorithms in healthcare.

Listen to the Episode Now


Ryan: I’m your host, Ryan Davies, and I’m hosting today’s topic, “Navigating the MedTech Maze, Predicting the Future, and Defying Startup Odds,” with Amir Hayeri. Amir. Thank you so much for joining us here today.

Amir: Thank you for having me

Ryan: A bit of a background for our listeners on Amir. He’s got quite a diverse background in developing novel medical software technology and using machine learning technology. Amir developed a technology to anticipate and prevent chronic disease complications. He’s presented this technology at a number of American Diabetes Association Scientific Summits. He is actively contributing to the science of glucose monitoring and how it can anticipate and prevent glucose-based conditions for patients with diabetes. Bio-conscious is the company. And that’d be a great place to start. You’re out, you’re out in Vancouver; I grew up out that way. Tell me a bit about the bio-conscious journey with your background and how it came to be.

Amir: Sure. Well, first, thanks for having me. So I come from, I guess, a long line of healthcare professionals. My family has been involved with health care for generations. And you know, I was raised in a family. My parents really like me to be involved with health care. So I kind of grew up in that household, and then I kind of feel like my generation really likes machines. So, my passion was with the rest of my generation. So, computer science and machine learning AI are what really excited me. So when I was done with my education, my kind of thought process was, well, ok. Now, it’s time to specialize. Do I want to go back to medical school? Do I want to go back to you now or continue on? Kind of my educational journey? I did have ambitions. I was just letting time go, and I’m not really doing something real; you know, academia is awesome. It’s great, but you do mostly research. And I wanted to hit the ground running. So the thought process was, is there a way for me to bring the two together? I knew enough about the biology in the background. I knew enough about what we could do with machine learning and computer science. So I had some projects, and one of my famous projects was replicating what your front desk nurse would do. But they’re awesome research projects. And one of the places where I think UBC really helped was kind of like drilling this into my head, like, you know, focusing on one condition and doing it really, well. So, a good friend of mine, who was a type one diabetic, always complained about the fact that he was late to class. He was managing, and he wore a Medtronic pump, I believe a Medtronic sensor, which contains a glucose monitor. He was very excited because he had just gotten a Fitbit watch back in 2016 and 2017. And me being who I am, I was wondering it was just like, oh, ok. Well, so I can take in all these three inputs, and maybe I can build a shitty algorithm that ties it all together. Then what we’ll do is use machine learning just because I knew we could do that to learn how he behaves. And maybe we can predict that because it depends on what type of machine learning you’re doing. But one of the first things that you would be able to do when you look at time series data is that when you have a really good model, you’ll be able to learn certain types of behavior without knowing exactly what’s impacting or affecting its behavior. And you can replicate that movement with the margin of accuracy, if you will. So I did that for him. it will take his heart rate variability from his Fitbit watch. It would take in his pump information, and his Dexcom sensor contains glucose monitoring sensor information. It would learn it for about 7 to 10 days, and then it would try to predict what’s going to happen an hour before it happens. This was revolutionary because up until that point, he was only able to respond to high-risk medical events or complications associated with his condition, which was type one diabetes. I don’t know how much you know about it, but there are different types of diabetes because of the spectrum: juvenile diabetes or type one diabetes, the type of diabetes that you’re born with.

Bio Conscious and the Journey

Amir: So your body loses its innate ability to manage your glucose levels, which is very important because if you have too much of it, you’ll die a slow, painful, gradual death, a little bit sooner than others. And if it falls, there’s a risk of immediate death. So you have to, on top of everything else you do within the daily kind of life, you also have to think like a pancreas, right? You have to like, oh, it’s high, it’s low. I need to do this. I need to do that. And so, using the prototype that I developed, he was able to manage better. So, he was able to anticipate events and then respond to them accordingly rather than waiting for his glucose to drop below a certain threshold and being alerted that it had dropped and then reacting to it. And there’s this really interesting thing about the one hour one, you know, 90 minutes or so horizon. if we tell users, for instance, something bad is going to happen today, like 23 hours from now, they usually don’t do anything in the daytime to do it; they’ll wait two hours or so, and then they only react to it within an hour. so it seems to be like it’s far enough where the user is not stressed. Still, it’s also close enough that they feel compelled to do something about it anyway. So he used his Ben really uses, enjoyed his sort of prototype, used it, and was interning at BC Children’s Hospital. So, he talked to the endocrinologist that he was working with. She invited me in, and we did a pilot study, potentially replicating the same thing 10 other times for 10 other teens and preteens. And that was awesome. Like, you know, this was one of those like high moments. The journey has had several ups and downs, and certainly, this was one of those ups. So I got invited, and we got started. We had two major companies providing devices for the study. We essentially replicated the same thing I built for Ben for 10 other kids. And we did it for 60 days, right? So 30 days of, you know, learning users’ behavior, fine-tuning the model if we have to, and then another 30 days where users who are using, type one Child, teens, and preteens that are using a pump, a continuous glucose monitor. and would be willing to wear a fitness wearable, which would essentially allow the machine to see different types of glucose behavior and get better and better and better at it. And we did it with the help of BC Children’s Hospital. Both companies donated devices, and it was awesome. We looked at the results, and the initial results were around 87%. Within the 1st 30 days, we’re able to bring this back to, if memory serves well, the lower nineties. And we were quite happy with it. And so we published the results at the American Diabetes Association, and that was essentially when I was kind of going to do this study and do the results was how kind of like bio conscious sort of morphed together. There’s a whole lot of backstory in terms of how we got there. But essentially, I thought this was good. So, you know, you asked me about the story of bio conscious. I thought, you know, that was a perfect name for it because we want to build a computer algorithm essentially that understands human metabolism. And you know, continuous glucose data is the perfect place to start for building such a technology because your entire body is effectively very similar to your car, except that one’s burning gas.

Your body is burning glucose. And the beauty of it is that our bodies are marvelous machines. They use glucose, the substance itself, in various ways, right? So we can talk about this whole spectrum of how your brain is the number one consumer of your glucose throughout the day. to the point where you get a cut on your hand that healing itself, how much scar tissue is left, And how fast it heals all of it is determined by your sensitivity to glucose. how your body metabolizes glucose. So that’s kind of like the story in a nutshell if that makes sense.

Challenges in Data Collection and Algorithm Design

Ryan: It’s amazing. So you were able to collect this data, which was, I’m assuming, scattered, and you know, you had data points to go off of, but you need to find predictors and commonalities through machine learning and data modeling. How can you collect that data, bridge that gap into an algorithm design, and bridge that gap into a product that seems like three entirely moving mountains, a kind of challenge between the three?

Amir: so well, it certainly wasn’t easy. So we had to do it,  in stages obviously. Originally, for the pilots and how we wanted to do it, we had to move really fast and quickly break things. Nobody really cared because obviously people, you know, patients and parents that signed up for the pilot project were told that this is just, you know, we’re cobbling a bunch of things together, may or may not work. Be careful, use common sense, et cetera, et cetera. We would get that data directly from the device itself at the time at the point of care where the patients use it in the long run, though, when we want to scale this sort of technology in later years. then obviously, we had other challenges. We noticed that initially when we got started, the regulation was at the time that the company that develops the device owns the data that it produces. It’s not the patient. It’s, in fact, the manufacturer gladly thought that within the first or second year that we got started, there was a massive change of regulation within the United States. The Congress turned around and said, no, the data belongs to the patients. The device companies are now only the custodians of the data. And suppose the patient, in fact, wants to share this information. In that case, the company is obligated to provide an infrastructure and pathway for this information to easily flow to where the user wants it to flow, which was great worked out. It’s really good for us. But when we try to collect the same type of data. So, for instance, your wearable data is not at the point of care, but rather, let’s grab it from the manufacturer’s servers. We noticed that we have these segregated databases online. Nobody really looks at them. They just sit and get overwritten. So, for instance, your phone records the number of steps you take per day. Your number of steps is like, what do you do? How is that going to help anybody? So, that sits on your Apple database. If you have a blood pressure cuff, that’s on Enron’s database; for instance, if you have a wearable, it’s on fit bits, sort of servers. And so bringing all of that information in a cohesive manner. So it’s nice and neat, and Mable was, in fact, very much a challenge. So the very first thing was like, OK, so it’s segregated, it’s on Apple’s device, it’s on, you know, manufacturer’s website, it’s on wearable manufacturer’s website. So we would grab it and then obviously pre-process it, normalize it, and put it in a place where you can now do a lot of iterative machine learning, try different techniques, and try different things. That was a challenge in and of itself. And then once we actually resolve that, then it is completely unrelated.

Amir: But the interesting challenge to resolve was that when you try to do machine learning algorithms that are trying to predict future behavior, You have to split your data early. You don’t want this to contaminate when you’re training or validating it because the algorithms are going to see it, and it’s going to try to predict the future, but you’ve already shown it to the algorithm. So, we call it future leakage. So we had to make sure and put environments together where this doesn’t happen because it’s a concern with all the models like you might get really good performance. Still, it’s because, when you’re training or trying to validate it, you’re actually doing it the wrong way. So once we resolved that, then the question was OK. This is awesome. But where is this data going to go? We’re going to analyze it. However, your care provider doesn’t review any of this information on any of these platforms. That information is hosted on your electronic medical record health record system. So now I open another can of worms. It’s like, OK, so if we want to get this data to the person that matters, which is your principal care provider, be it your endocrinologist, your non-diabetes, expert family doctor, that person needs to see how you’re doing. That’s the whole point of what we’re doing, right? So, trying to get it to that ecosystem was another challenge. Everybody says they’re either HL seven or FHIR compliant. But when you go actually deploy, you, like, no, absolutely not. They have their own little thing that you have to build, and that takes multiple months. So, that was also a huge challenge to try to resolve.

Amir: And all of these data points are in a different format, different types of looking at it. The same thing with electronic medical records and health record systems, where the physician keeps notes, differs from physician to physician often, right? The reason why I’m sharing all of this is because that’s awesome. You resolve all of those back-end challenges for the data to flow and for something to happen. Then, when we fully resolved it, the question was, OK, how do we display it? Because at that point, we had, you know, a certain data flow, we could get real-time information that the patient is producing at home and predict high-risk complications ahead of time. But this information needs to flow somewhere; somebody needs to do something about it, and it needs to flow up to the clinic. And so when we finally got connected with the MIHR, we noticed that they can’t display the data that we want them to, you know, primary care provider. And that was kind of like, oh crap. We have to do all of this now, and we’re just like, ok, that’s the mission. When we do this or golden, we’re going to exit. This is going to be awesome. We did it, and it’s like, we need to display this somehow to the user, and we couldn’t. So that was like the theme of this whole journey, which has been: you solve one problem, you find 10 others, solve another 10 others. And it’s an ongoing thing. But gladly, because of the team that I was fortunate enough to have, that team has morphed and shaped into different things and evolved. But today, I’m very happy with how far we’ve come. And it’s just been because I was crazy, and the team was, you know, the team I had around me was able to be open to my crazy ideas and be agile enough and fast enough for us to outmaneuver some of the major sorts of competitors that we encountered along the way. But today, because of their efforts, we have a system that I’m very proud of. At the end of the day, it comes down to keeping people healthier for a longer period of time.

The way we do that is we try to learn users’ behavior and predict near-term high-risk medical events, and we do this for people who need it. So, for instance, patients with type one diabetes and patients with type two diabetes intensive care that uses insulin. And hopefully, you know, if the algorithm keeps seeing different types of glucose behavior, which we have preliminary data on today, it will learn more and more about how our bodies function. One of the things that diabetes has done, unfortunately, has skewed our understanding of glucose metabolism. And that’s like one of the things that we learned along the way. is that the technology that we’re building today has implications in predicting near-term complications for patients with type one and type two diabetes. Type two, but one day, it can actually predict different types of glucose behavior just by looking at how your body deals with your different types of actions on a daily basis. Is it eating? Is it doing any activity, is it resting? And when we look at it, it’s fascinating: your body has a different profile of metabolizing glucose during sleep. There’s a different way of metabolizing it when you’re active, a different way of way of using it when you’re stressed. And so in type one and type two, the machine has learned or has become smart enough that the algorithm is smart enough to be able to pick differences. So, we cluster or create different buckets for different patients. And that makes me very hopeful. But anyway, I don’t know if that answered your question. I ramble. Let me know.

Partnerships and Challenges with Competitors

Ryan: No, I think that was great. And there was, like, I made notes of like 10 other questions that came out of this. And I know we’ve got a limited time here. But one of the things I want to ask you about is the challenges that come with partnering. Because you’ve got a whole bunch of different access that you need, you need partners to help you with this. You were talking about smartwatches, but also like access to things and whatnot. And then you might have some partners who are in the space going well if you can do it, so can we, and we’re bigger than you. And I might have more money than you. So I don’t really need you anymore. How do you walk that line?

Amir: Listen, I’ve lived through that like three times already. The first time was devastating. I was almost going to change careers. So this is what happened when we did our very early study, our initial study, people like this bear in mind. This was, like, 2016, 2017, that we did it. the idea was novel. It was crazy. It was like, we’re coming out of academia, and it’s just like we’re hot and bothered, and we want to do something. we did something, and it worked. then we took it to BC Children’s Hospital. They like it. They saw the potential. They’re like, we know it’s cobbled together, but it could be beneficial. So they shaped it a little bit. And then, we requested for devices, and these companies were like, oh my God, this is like they read the protocol, and they’re like, this is awesome. Here you go, a bunch of devices. and we use it. But what that meant was we were contractually obligated to share the results with them. So be careful if you get devices from these companies; make sure to read the contract carefully. But that’s what it means. We did this study, and the results were good. But unfortunately, despite what they told us in school and business courses, these companies are firstly incentivized to try to replicate you. If they can’t replicate you, they’ll try to buy you. But if they can replicate your technology or get close enough, you’re done. So, one of those companies that were partners just decided to have this love-hate relationship developed with us where they loved what we were doing. Still, they hated us for doing it, and they thought to themselves, they have a room full of engineers from Boston, MIT, and Harvard. Was it this buzz out of Vancouver? so they obviously started doing it. That was devastating. They stopped working, they didn’t stop working with us, but they kind of like, ok, stop paying attention to us. All of a sudden, you’re like, you’re not that important. And then they’re so certain that they’re going to do this in a very short amount of time. They actually decided to do a press release on it. And that was like shit. The person that we tried the prototype on actually sent it to my family and me, and I still remember it to this day and get goosebumps when I talk about it. We were driving back from Alberta. I don’t know why we were there, but my friend sent it to me on Facebook.

Pivotal Moments and User Interaction

Amir: The press release came up, and I read it, and I was just like pulled over and was like, I can’t drive. My dad took over what happened and just like, yeah, so this happened, and I was like, oh, I’m so sorry. And I was like, yeah, me too. And, like, didn’t say a word until we got back. That was an awkward, long, silent drive. But when we got back home, and I kind of like shook it over, I realized that there are two things that we have that they don’t, and one is being fast and agile, and at the end of the day, they are trying to copy my idea. I can bob and weave and move really fast and try to outmaneuver them. if I’m smart enough, so put a lot of pressure on it, and that kind of pressure is, you know, talk about loneliness and everything. That kind of pressure only does two things. One of the things is that it either breaks you or pushes you to reach really deep down and find something you didn’t know you had. And so far, I want to feel like we’re still around. So, I’ve been able to persevere despite the pressure. So, I thought to myself at that point that every diabetic app that I’ve come across has been very serious, and it’s talking about something extremely serious that, you know, oh, your glucose has dropped. Oh, this has happened like it’s just got this negative, you know, feel it. What if we did something really simple beyond machine learning? What if we made this app that’s not able to predict future events for patients? But it’s also a little bit sassy and gives you a bit of it like, oh, is it a bird? Is it a plane, not, it’s your blood glucose? Just, you know, corny corky, but it’s stupid enough that it’s funny, and it makes you laugh, and obviously, we found a bunch of them and put it on random, and we said, let’s just outmaneuver these companies. 

Amir: They’re going to try to replicate it. Let’s have a product, it works, we’ll just, you know, put it on the app store, and it was serendipitous because right after that, within like a few weeks or so, a mother with a type one child found it took it to one of these, you know, hidden Facebook groups for moms with type one. people started downloading it, and it was funny because we built a little code the guys built because I wanted to really keep track of people using its user experience. How is it? So, they attached it to Slack. So every time we had a new account successfully made, it would go ping. So, when it’s not when you’re not getting accounts, I would be notified, or they would be notified, something is broken or something. So anyway, it was going bing every now and then. So I was happy one morning. I woke up, and I was just like, bing bing, bing, bing bing. I was like, damn it. The code probably broke. We have to redo this. And I’m like emailing my co-founder Ricardo. It’s just like Ricardo; it’s just that this thing is going off like it’s just, what is wrong with it? And he goes in. Obviously, he is in the zone out of it, doing his thing like OK, great. What is it? It’s like, no, actually, we’re getting a ton of users old. Let me check the database. The number of users is going up, and it was really good. So, within four months, it was mid-2018, give or take. We had 15,000 active user accounts on our awesome database. Tons of users, right? We have multiple functionalities, follower, uploader, et cetera, et cetera for patients and followers. Still, we had a bunch of users. It was really good, and during this period, what we noticed was that that type of like give people a little bit of sass like, you know, just offer something a little bit out of the ordinary if you will, kind of get people to, they had a binary reaction to it, right? They would either love it or they freaking hate it. But regardless of that, if they love it, they’ll take it to Facebook and post about it. If they hated it, they’ll reach out to us and tell us why they hated it. For instance, we had this around the time when Trump was running. Make America great. Again. That’s famous. The famous phrase is just like sending it in, like making your blood sugar management great again, or like we had different variations. I don’t remember verbatim. But it got to a point where it just like it polarized so many people, they’ll reach out to us just like, OK, you’re doing this, you’re doing that, et cetera, et cetera. Of course, we loved it because that meant an interaction with the user, collecting it, right? Information like you have to take these calls, you know, regardless of good or bad, you want to talk to people who are using it. And it was really fun. Unfortunately, that led to a situation where we have a ton of users, and we’re trying different types of algorithms. So, at any moment in time when we had a new user, 72 different machine learning models with different horizons, different objectives, and different architectures would start running, and these would run on server farms, right? Like Amazon Aws. So they would run and, and so 15,000, you do the math, like 15,000 users active like 72 models per et cetera. Like, I don’t know how many that is. But what I do know, though, is that our server time came to, like $ 7,000- $8000 per month, and for server time, which is relatively cheap, that was high around the time. It was like, oh, that’s like another full-time developer for the team. And I was spending my own savings and, you know, I had a trademark that was bought, which is really good, but story for another time.

Struggles and Triumphs in Funding

Amir: So, I was spending my own money trying not to attract funding and investment up until that point. But then it got to a point where these like bills were high, and if you don’t pay Amazon, I think it got to a point where we haven’t paid them, like, if you don’t pay them for three months, I think they cut your service right away. So I think we had two months, and we’re just paying them a month. So it doesn’t turn into three types of things. Oh my God, it was awful. So, I think I processed the last payroll, and I sat down with my two developers at the time. I think Ricardo and Savon that we had, we had somebody else working with us as well, brilliant machine learning scientists. But anyways, I sat down, and I was like, guys, I just processed your last payroll, and I don’t have, like, literally, I only had like 50 bucks or something left in the company account. Kind of just like, that’s all I have. And if I can’t attract funding in the next 30 days or so, count this as your 30-day notice. And it was awful because we’ve come this far. We’ve seen, oh my God, we’re almost there. Look at this, actually. Like I was showing the guys, I still have to this day. I took Facebook pictures of the reviews back then. And when I present, I refer to them because they’re just awesome, right? But, you know, it got to that point, it got to the point where it’s just like it’s done. I can’t, I’ve spent all of my money. I borrowed money from family, like everybody, just like nobody was willing to give me a dime. It’s literally just like going away. And so I kind of was like, shit, what can I do?  I have 30 days just investors like whoever is an investor or pretending to be one, which, you know, most of the time people are pretending to invest or not really invest anyways. I’m going to meet with all of them. And so that’s what I did intensively for like 30 days or so. I was doing it before, but not as intensely as I did it. So I had like 5 or 6 meetings with people. It was really tough on my mental health. I really just wanted to sleep because it was like the same type of questions, different people. Some would get your answer. Some wouldn’t. And then you can’t get pissed off or emotional about it. Like, you need to understand why they are not getting it, leave the room, and make changes so that it doesn’t come up again. And, doing it intensively, there’s only so much time you can, like, operate on under stress for a really long time. It takes its toll, and it was certainly taking its toll. I remember I was at the office that we had. I was by the front desk, and I had me, like, last, sort of meeting. I was like Wednesday or Thursday, and I was talking to the front desk, the lady at the front desk, Emily. I just want to go home and take a nap, and she was, like, go home and take a nap. You look awful.

Amir: And I kind of was just like, I know, but there’s this event I want to go to which I, knew the owner, there’s this conference that, one of my friends runs, in town, and he was like, I want to update it. So, why don’t you come present? There might be some people with money there. I’m sure he was pitying me. I’m 100% certain that he was. Anyway, I knew the guy, and I didn’t want to let him down, but I really wanted to go home and take a nap.

Have I gone home that day and taken a nap? It would have had a completely different ending to this story. So I went there despite all of it; I’m just saying go there, do this. And back then, I remember I kept saying to myself, you know, you miss 100% of shots you don’t take, right? I can’t afford it. There are periods of time when you can afford it. You’re like, ok, you know what? This is not a high priority. I’d rather focus on a better priority. But at that point in my life, I was just so goddamn low, but I didn’t have that opportunity. I would pitch like I was going to the US. I can attract investors and money from Seattle, obviously, because of the proximity. and I found myself in a situation where I was talking to a TSA agent about getting a device on and getting my app. And I kind of realized that he’s just asking me, like, why is your business a pleasure? I don’t know. I was just ready to go. How far? That was the thing. And so anyways, at the end of it all, I went to that event I presented, I was at a point where I had like 100, and something slides, and the guy was like dude, you’re just five slides. What the hell? And I, like, listen, it’s five, but like, I want to bring it up if they’re questions and they’re like, fine. That’s ok. You know, so I got on stage, started the tape, you know, did my thing.

Some questions came up, and it was so funny because one of the judges there, who is actually a seed investor, had heard my pitch probably three times before that specific event. The video is still on YouTube. You can find it. It’s at Cambridge House. That person saw me and heard me before we were like our office back then, and his office was within close proximity; ran into each other at the elevator and whatnot. And it wasn’t up until that moment where he was forced to sit on stage and grade or, you know, find the winner between, if they had a competition thing. It wasn’t until that point that he was forced to sit down and listen to what I was actually telling him.

Amir: I swear to God, it was the same message. It was the same thing delivered four times, and the fourth time was the charm, despite what they said. And, you know, one thing he said is just like it merits a further look. It’s like, ok, well, that’s a good change. answer the questions and everything, and then thank you very much. Get off the stage and, you know, I emptied my pockets because I had like a million things in there, put everything back into my pockets, and I’m just like, oh, I’m missing the flash, the thumb drive that I brought with the slides. So I’m going around the back, grabbing my little thumb drive, and then I’m just walking back, and they’re doing their thing, obviously. And I’m just going home, I want to take a nap before this whole thing, and I like, ok, I’m on a mission. Yeah, just got out of my way, and I saw the front that there was a girl behind the stage. She’s like, where are you going? And I’m like, I’m going home. I just take a nap. I’m really tired. And she’s like, no, you need to go back on stage. They’re asking for you to come up there. And I’m like, why did I do my thing? And it’s just like, no, you won. I’m like, really? It was a competition. So I ran back up, having no idea. I swear to God. The guy says like, oh, you won, blah, blah, blah, something I don’t remember. And I like what seriously and people laugh, and I’m like, you know, trying to be funny or something, but at the time, I had no clue what the guy was saying. It was so unimportant, you know; I was just like, ok, today is done. I didn’t manage to do anything.

I had to go home, and I needed to take a nap. But a lady comes up after the event. we talk, and obviously, she becomes our lead investor. I got very, very lucky because of that we have been in several situations where, you know, almost everything else. There was like the pandemic. There was this: that the markets were bad. We got to a point where we ran out of money several times. It was the fact that our lead investor was both in a position to help the company but also believed in the vision enough to do so, despite the fact that, hey, you know what, this doesn’t look good. Still, it’s really important for anybody who’s out there who wants to find a lead, you and the lead need to align.

Amir: in terms of what the objective of the product is, if the objective is to let’s make a bang for our buck, that’s an objective. But at some point, you’re going to start diverging from one another. And the lead investor needs to see the value in the technology and the mission. And in multiple times like, listen, we’re going to do this, we’re going to do that. But at the end of it, we have this option as a last resort, or we can do that. And she’s like, OK, no, you know what? It’s fine, don’t do the last resort. I’ll do this, I’ll do that. And that was really important because otherwise, we’d have a completely different sort of story. But I noticed, for instance, you know, having a lead investor that happens to be a female also is something that people respond to in a very interesting manner because most of the time, your lead investor is either a man or a VC or an agency. Oftentimes, not always, when you do have a lead investor, that’s out of the ordinary. If you will, that makes others pause and, kind of like, have a different reaction than they would otherwise, and I can talk about this later. But anyway, I’m rambling again. Is any of this helpful?


Ryan: No, this is exactly. This is absolutely perfect. I mean, out of this, we’ve already got two things. First, we need to do part two of the rest of the story here because that’s there. And secondly, delving into a strategic partnership with a lead investor is a unique aspect in itself. There’s a wealth of valuable information for many to gain from this.

Amir, I don’t believe our conversation is complete. This might be a brief pause in our podcast journey, but we’ll need you to return and continue sharing your experiences. We’ve only just begun to explore the intricacies of the medical tech space and the art of navigating the medical tech maze. How to overcome startup challenges, avoid being one of the nine out of ten that fail, and deal with the associated anxieties – these are topics that deserve a more in-depth exploration. I envision the possibility of creating an entire series based on your insights, my friend. Hopefully, we’ll have the opportunity to do so. With that, I acknowledge that it’s time to hit pause as we’ve reached the end. A heartfelt thank you, Amir, for an incredible podcast session on navigating the MedTech Maze, predicting the future and defying startup odds. As I mentioned earlier, it feels like we’ve just scratched the surface, and I eagerly anticipate diving deeper into these subjects in future episodes. Thank you so much for sharing your time and expertise today. We’ll be reconnecting for another insightful conversation, that’s for sure.

Amir: I’m glad you find my ramblings helpful.

Ryan: Everyone’s going to find them helpful and entertaining and wait for the next episode to continue on because this was just fantastic; it’s a real story, right? There are lessons to be learned from here. It’s truth, there’s heart and soul, there’s ups and downs. And I said, we’re just touching the surface of those, and I can’t wait to get through more of your stories as we continue on here. I want to thank our listeners as well. We can’t do what we do without you. So, until we meet again with another amazing TBR episode, this has been one of my favorites. I am your host, Ryan Davies. Thanks, everybody. Take care. We’ll talk soon. Thanks.

About Our Host and Guest

Director of Marketing – Ekwa.Tech & Ekwa Marketing
Read More

“The challenge to resolve was that when you try to do machine learning algorithms that are trying to predict future behavior, You have to split, your data early, and you don’t want this to contaminate when you’re training or validating it because the algorithm is going to see it and it’s going to try to predict the future, but you’ve already shown it.”

– Amir Hayeri –