Bản chép lời đầy đủ của The Long Build: How EyeQue Grew a Decade Without Venture Backing

that use our system. So they have to pay a monthly fee to get access to our platform. And then we also charge for per prescription. There's a fee associated with that. And we also have to pay the doctor. there's a number of flows going on, it's those are the probably the four major streams. ⁓ we have done some licensing some of our apps. That's a ⁓ stream in the future, but we Amanda: Mm-hmm. John Serri: not grown enough stream very much at this time. Amanda: And we are the Madison Advisory AI podcast. So I will dive into AI eventually down the stream of thought here. But before we go into that, what made you rethink the business model as it was operating and decide to change or veer into perhaps a different direction? John Serri: Yeah. Yeah, that's a good question. ⁓ is just, ⁓ has to do with the people in the company, including myself. We're ⁓ ⁓ orientation, so we like to play with things. We like to explore new ideas. ⁓ Over the years, the company's generated 35 patents. About 20 or so have been granted so far. We have a number of others in process. So one of the things that made us do more things is innovation. ⁓ Also, there was a desire, ⁓ you know, eyes we see in a binocular fashion, This is a monocular device. You test one eye at a time. So we were driven by creating binocular type devices, which we have created for some in-office type of purposes. We actually have binocular versions of devices. When you go from a monocular to a binocular, it sounds, oh, you just put two of them together. It's not that easy. So you have to worry about things like the pupillary distance. And it gets into lots of human factors engineering that you don't have with a monocular device. So that was a drive. Now, one of the things that we've kept as a constant is to use a low cost vehicle for the user. In other words, we don't have a special display or a special computer that we sell. You just can use your smartphone. Now, part of the work too, was evolving to be able to use applications that... across different operating systems. For example, we support iOS, we support Android. That's a lot of work needed to support both. And then dealing with the App Store is one of the things that we evolved to is moving to pure web apps. So in other words, it's just based on going to a website URL rather than going to the App Store and downloading something. So we've evolved in that direction as well. So we were driven by curiosity and also we believe that going to more devices would expand our business as well. You know, there's a quest for accuracy and other things. And of course, the other factor was the regulatory nature of vision and other things. So we moved to a digital platform. When I started the company, I had no concept of creating a telemedia. Amanda: Mm-hmm. John Serri: medicine platform. But after a couple years it became pretty obvious that that was the way to go. So those are some of the factors. Amanda: What made it obvious that going to a telemedicine platform was the way to go? John Serri: Well, you want to provide good service to people and you want to grow the business. You want to work within the traditional avenues. A big factor was COVID. Okay, so COVID came 2020 changed everything. And it, more and more people move to telemedicine, digital kind of virtual health, which by the way, I believe is inevitable. Any investors are thinking about what the future is. You're going to have less doctors. manually doing things in offices and more and more stuff will be done online and through AI as we can probably get into later. So we wanted to work in that direction towards providing a good service for people. Amanda: Okay. How has the company changed in the past two years? John Serri: ⁓ Well, we've implemented what's called, the term is VRI. It's a virtual evaluation of refraction using artificial intelligence. So basically using the same device, we've built a web app that makes ⁓ the testing much faster because it uses your previous results. to tell you where to go next. With device like this, again, I'm getting a little technical here, you have some slits that you have to rotate on if you can see them, but you need to measure what's called a stigmatism. So that's by changing the angle at which you perceive some lines and you make some corrections on them. So if someone doesn't have a stigmatism, which is a large fraction of population, there's no need to make all of the Amanda: Okay. John Serri: measurements. So once you determine there's no stigmatism, okay stop. You don't need to go and make people test more. So we've been able to speed up the tests by about a factor of three overall with this and the accuracy is quite good with it. So that's one of the directions that we've gone in and we've expanded the telemedicine service both through the software, streamlining it. So those have been the ⁓ changes over the last ⁓ years. ⁓ haven't ⁓ in brand new, ⁓ ⁓ hardware technology at this point. It's mainly been, ⁓ have a business model, we have a device, we know the technology works. whether we do it in the future. We're still grappling with that issue, of course. But that's sort of where we've been. We've also, in the last three years, we went to the B2B model with the optical stores. And we created a very nice device, which is called Insight Pro, which is ⁓ an in-house testing platform where you have a, we provide the tablets and smartphone and a ⁓ binocular viewer and people can take a battery of tests with that. It goes directly into our platforms integrated directly in. So there's a lot of things that we're working on. Amanda: How has AI, if at all, helped in user error of using the device? John Serri: Yeah. ⁓ In terms of user error, we get somewhat better performance because of what I just described. We have an algorithm that eliminates the need to do lots of testing. ⁓ that's been there, but the potential for using AI is vast and we are going to harness it. We're beginning to now, but the

benefits with AI are... Orders of magnitude kind of benefits, I would say. Yeah. Amanda: So let's dive into that because there are benefits it seems like on both sides of the business on improving the technology and how it's used, but then also on the business model itself. So on the technology, how are you utilizing AI today? John Serri: Well, in terms of just the speed of software development, as I said, we're moving more and more to web apps. We can get stuff developed with much fewer people. In fact, a handful of people can do the work of what before was maybe 10 to 15 persons. Okay, that's a big deal. So being able to do that is... It's big cost saver and a time saver. So I suspect that will get better. we are, as the company goes forward, we're going to be using more more AI in our implementations. So for a cogent, very good. Amanda: Where does AI augment the business versus replace? I guess in the developers, it replaces personnel. John Serri: That's a case of replacement. mean, in a way, augmenting replacement, there's not a fine line when I think about it. But the other area is in customer support. Amanda: Yeah. John Serri: With our testing, it's a subjective test, meaning the user has to look at a pattern that they see on the phone through the device and make a judgment, then enter that. So there's a number of steps required. This is a human intervention. It requires effort on the part of the person. That subjective nature of testing is not going to change. The best person to tell how well your eyes are working is you. It's your brain. It's what you see. So that doesn't change, but what? But what happens is a lot of people, it's not a trivial test. That's one of the challenges that as we move forward we want to make the test easier and easier. However we can do it, there's lots of things we can do. So one way we can do that is through more active voice prompting. the test, for example, will give you nice verbal commands in whatever language you like. Right now we have some verbal commands, but I mean the app talks to you, but now you talk to the app and it does things. So you can ask the question, it does things. So building that type of interface would be very helpful if we'd done in different languages. Also, you can start to look at the results that the person is getting in testing, looking at the statistical consistency. If the consistency is low, there's a problem. You need to talk to them. You need to say, why don't you talk to support? Then there's another avenue, there's another piece to this, which is the support piece itself. So people... provide support, we have users, some users take ⁓ time to welcome through. We have people with different skills, different levels of patience. In other words, patience as in the ability to stick to a task and hope for the best. So that takes time on a part of our support staff. Through AI, we can reduce that and we can begin to, you know, we're looking at introducing chatbots and other things. We have some activities going on in that area today. In fact, we're beginning to move that direction. So I think that too will be... So, so have AI interacting with the user directly to help them through the test process. That's for sure. And there's, you know, I can, there's other things that are out there as well. Amanda: How do you envision the company changing in the next two years? I know you've gone through an evolution now. When you look at the growth coming up, what do you see? John Serri: Well, yeah, know, it's, technology clearly works. Now it becomes, what do you do? Well, you do this, in my simple thinking, there's just two paths you take and you kind of want to take them in parallel if possible. Amanda: you John Serri: The first path is to continue to improve the product offering. And as I said, we're going to be able to do that through implementation of AI in a number of ways which I've covered. the other thing is to get out there. Okay. Directed consumer business is not easy. it takes a lot of money. You're competing against, ⁓ avenues. when people ⁓ look us ⁓ they're dealing companies that are not ⁓ quite doing IQ, in fact, no one does exactly what IQ does, but they're similar, ⁓ but they huge budgets. that when, ⁓ when ⁓ go search engine opt-in, and stuff like that, ⁓ the company would have to pump in a lot more money to do that. Okay, so we are looking to... basically get more capital. That can be done in different ways. One is to get some direct equity investment. We kind of like that idea. The other one is that the technology and IQ can live well in a bigger outfit that does similar things, can act as a competitive edge and It's a. of marketing vehicle that would be massively or to the magnitude greater than the market access that IQ has today. So we're looking at both of those. The goal is when you take a step back, technology really helps people. the trends are in the direction for this technology. if you look at the number of ophthalmologists in the United States, that number decreasing and the number of people going into the difficult process of becoming an ophthalmologist after their their receives are empty they have to go into the residency those numbers are actually lower than they were a few years ago. Less people are going into the field. Same is true in ophthalmology. On the other hand what's happening is that the population Amanda: Mm-mm. John Serri: Per person, the need for eye care is going up. There's two factors that, two big factors that are played here is what's labeled as the myopia epidemic, which is that more and more people are developing nearsightedness and this happening at a very early age. Okay, a lot of that's attributed to the use of digital devices like smartphones and people spending less time outdoors. So you have this myopia epidemic and it's huge in China, for example. Something like 90 % of kids by the time they finish high school have moderate to severe myopia, which means they have

to wear substantial glasses or get a lay-sick or it's a significant thing. The other thing, there's a genetic component to it, okay? But. Amanda: Okay. Is that genetic at all or is it really technology based environment? John Serri: There seems to be, and there is a strong environmental component. There used to be much less myopia years ago. What's the difference? People used to spend a lot more time outdoors. People didn't spend half their day playing video games at six inches from their eyes. So those two factors, there's a big environmental component for myopia. That won't go away. Amanda: All John Serri: Maybe you might find some kind of drug, some kind of medicine that can be used to reduce it. But basically what myopia is, the eyeball gets elongated. And so the focus that your eye lens does actually happens before it hits your retina, which is where the imaging should be done. So this is the problem with it. The other big problem is the aging population. After age 45, everybody has the need for glasses, with very few exceptions. And this is something called presbyopia. It's called... your eye no longer is able to focus up close. It stiffens. The actual lens stiffens. The muscles that control the focusing ability of the eye lens decreases with age. So that's why you get people wearing ⁓ reading glasses, right? They're called readers. So they act to magnify. ⁓ That's growing fast. ⁓ then you have the serious diseases ⁓ glaucoma, ⁓ retinopathy, ⁓ of these things increase with age, Macular degeneration. So there's gonna be more more need to ⁓ diagnose ⁓ And can diagnose very effectively using online methods. So the market is huge, huge for this. Even today, ⁓ Large fractions of the US population do not get an opportunity for good eye care. There's just not enough of it around. And insurance is tough and other things. These technologies reduce the barrier. They reduce the barrier with respect to price. They open up access to people. Almost everybody has a smartphone. Amanda: Mm-hmm. Right. John Serri: This was one of the premises we based the company on. This technology would eventually become very important. So to me it's an inability that it will happen. Amanda: Yeah, so the need really speaks for itself with one technology, the usage of it really expounding, especially the past decade or more, and then the aging population. the way you see the market and you see the market growing so substantially, what do you weigh when you're deciding between whether you raise equity or find a partner? John Serri: Yeah. Looking ⁓ both, ⁓ gets into all sorts of ⁓ Some of them are ⁓ related what ⁓ people to do with their life. In other words, founder, ⁓ do want to continue? Do they want to retire? ⁓ There's element in it. ⁓ There's a strength in the commitment to the technology, the commitment to do good. Some people may prefer to cash out, others may prefer to continue to push the envelope on this. So those are the considerations that come into play with something like this. And it's not a simple answer. It's not like, ⁓ because if you fell to a good strategic acquirer that understands the potential of the technology, they can take this thing. Amanda: and No, it's never a simple answer. John Serri: They have the connections with the regulatory, they have the international connections, both business and regulatory. They can take this a lot further than a small company can take it. And the net effect is there'll be much more good done for humanity that way than trying to just build up on your own. So those are the issues that surround that type of decision. Amanda: Right, there are pros and cons and one avenue may make the technology even more accessible and wider spread. And if that's the ultimate goal that could be one to pursue. When you think about investors in either lane, what do you think they really understand about the business and what is most misunderstood? John Serri: That's right. Most investors are looking to make money as I see it. They want to return on investment. They want to know that the technology can work and it doesn't have any kind of regulatory snafus or other things. People generally want to return quickly. So I think those are the factors that drive investors. A lot of the people I've talked to over the years, very few understand the technology. Very few even want to understand the technology. I've seen that. So someone to take on IQ, you need to... understand the technology and you need to feel the need for it. So you need someone probably that has a background in digital health, public health, ⁓ optics, eyeglasses to really appreciate this. Amanda: Mm-hmm. what do you think the company needs to prove itself ⁓ before go down any ⁓ sort avenue? John Serri: Well, ⁓ proving it now. ⁓ is not like a pre-revenue company. It's not like a ⁓ startup. ⁓ a lot of ⁓ practical ⁓ that's been developed. ⁓ there's a market there. So the company is at a good position to look for these sorts of ⁓ possibilities right now. So that's kind of my vague answer to that. Amanda: If you were starting IQ today and in this capital environment and then with this AI environment, what would you do differently? John Serri: ⁓ Boy, that's a great question. That's a really hard question to answer. You know, it's so hard to predict the future. It's so hard to be a Monday morning quarterback. ⁓ I started the company after many years working in aerospace business, working in government, working in physics, those type of areas. ⁓ I personally started this with a friend who was in the eyeglass business. In fact, he was the founder of one of the most successful online eyeglass companies in the world. That's a company called ZeniOptical. His name was Tibor Laksay. And I came in and after he approached me and says, could we do something to test glasses online? And you know, I said. sounds like an interesting idea. So I came into it mainly as a technologist. I do wear eyeglasses, I was not so much... ⁓ I was the

arms and legs, so to speak, to make the technology work. And Tibor was the true visionary who saw this. He saw this way before anybody else. He saw this back in the early 2000s, how... eventually eye testing and more more telemedicine would occur. was he was clearly well first in that in the beginning. So what I know today, knowing where AI is going, I would start a company with an AI focus. When we started the company, The focus, the technology focus was cloud services, right? That's where, and so we set up basically a cloud service for IQ. Now we would set up a cloud service but built on a heavy AI platform. That clearly would be the way to do it. You still need, now here's the point though, AI can do certain things. So AI can help you in design and other things, but you still need to build stuff. You still need to build hardware, know, little materials and focusing elements and other things. And you still need to have humans in a loop. We have eyes, we test eyes. So there's going to always be a, simply speaking, a hardware component to this type of technology. Amanda: Mm-hmm. John Serri: Refraction is the art of bending light. Refraction means the bending of light. You can't bend light just for pure AI software. No, you need a physical device to do that. But AI will help with generating optimal designs for which to go build devices. So that is true. So, but there's going to be an element of hardware always through any kind of medical testing, I believe. Yeah. Amanda: Right. I think that's a good point because we often talk about how much AI can do and how fast it's moving. But there in all AI uses, there's a human element and having the human in the loop, especially when it comes to health care is essential. And I think you make a great point there. John Serri: It really is. There's so much ⁓ AI hype out there. it's created the potential for the technology is tremendous. The dangers in the technology are tremendous as well. ⁓ so AI, you know, I spend personally time now on AI safety regulation. ⁓ I'm looking, you know, what's the implication of the socioeconomic implications of these systems? Again, once it plays into with less jobs around for people, you're going to have people doing more and more at home diagnostics. again, that plays into where IQ is headed. But in general, there's a lot of hype and the trouble with it is that people expect results too quickly. Okay. So you have people going out and saying, ⁓ there won't be any jobs in a year. And, or, you know, there's going to be massive, massive changes to the economy. It's not going to happen that fast. It's going to happen, but it's going to be longer. So I think investors need to kind of keep that in mind. If they're being hyped by someone about, we can get this thing built in three months and, know, automatically generate code and all, just have to do five, you know, state what we want and all. Amanda: Okay. John Serri: It isn't that easy, but in the long run the potential is huge. It's awesome. Amanda: Yeah, I think I agree with you. think the potential is there and that's the ending. Ending our conversation today is like on a great note that the potential is huge and I really thank you for this thoughtful discussion and getting into the tension between technology and capital and the how much ⁓ IQ itself has the potential to grow and perhaps find capital to grow or the right strategic partner. This has been a great discussion. Thank you. John Serri: Amanda's pleasure.