AI is an aid to the doctor, not a replacement

It's my pleasure. Thanks for inviting me. I I appreciate it.

well I think that what's important about that is I I didn't reali I'm not sure I intended to do this. So but there was an unmet need in industry, something that I had been working on for a while and it and it it it kept coming up, that these issues and challenges that our customers had kept coming up. And it's the kind of innovation that is not easy to execute or effect in a large organization. there are a lot of standard there there there there is a a kind of inertia in industry and and and so as a as a vendor, you know, you have the products that you want to sell and you have the support that you can offer. But if there's a if there's a gap there, there just aren't the structures and mechanisms in in a big company to to you know to find to allocate the resources and then to pursue those improvements. And so I I just felt like I really was just driven by the by the the desire to solve the problem, honestly. That really was the, you know, the the driving force. so that's what I set out to do.

Well, we we are focused on clinical trials. So we're we have training and technology to support processes in clinical research, specifically in the immunodermatology therapeutic area. and so some of our our products, well, one of our products is a pretty standard tool. It exists in a lot of areas, but Th this is all around the way that you evaluate a drug. So when you're testing a drug, either the EMA or FDA or in you know the the in in Japan, the PDM PMDA, you know, whatever regulatory body there is, there are measurements that you have to use in order to prove efficacy. So you're you're You're trying to test the drug and see if or or the you know the machine that it could be a device. and you're trying to test that and and really get an understanding. And so you have to set a benchmark for the severity of the disease and then measure in a placebo-controlled study. You want to measure, does the drug make people improve more than nothing or a placebo? and so that's w what we focus on. But there's a lot of kind of Imprecision in that process. It's a pretty subjective in a lot of ways. And so you're asking physicians, as they're called raiders or you know graders, in these research studies to assess this skin disease, inflammation, and severity of disease. and so that's where we fit in is we started by creating expert training. And then that training really becomes the basis of the software and the applications that we develop. right. You it's really important, you know, to understand the problem, right? And to understand how what the existing solutions are. So, how are these problems being approached currently before you try to affect change or affecting?

Well, I I do I think it's the basis of it it the basis of what we build is the the expert guidance, right? So in other words, we're we're looking at not what is easy for us to build as a technology or what would you know what sort of follows a logic. you have to have kind of two parts to this, right? You have to incorporate some of the subjective measurement, some of the subjectivity that is inherent in the process. And you have to really pay attention to that. So, what we do is we started by having a really deep understanding of what the measurements are, and they vary across j therapeutic area. I mean, across the therapeutic area, different indications are measured very differently, and the instruments that you use to measure them are very different. And so, you know, what what Technology likes to do very often, B2B technology, is they like to have one thing that is repeated and can be reused for you know for multiple things. In this particular space, that doesn't really work because the measurements are designed differently, they have different parameters, they have different requirements, the calculus is different for each one. And so you really have to have built a foundation, a framework. and then on top of that framework you can find specialized solutions for that that address the specific needs of each indication. I may be I mean, this may make no sense to you. I d I don't know. It's a very esoteric kind of thing, but

Yeah, and well it's not a one size fits all. Yeah, right. Yeah. Well and and but the other thing is is that because it there are different levels of so all of these diseases are at different phases of development, right? So something like psoriasis or atopic dermatitis, which people, you know, commonly would call eczema, atopic dermatitis psoriasis.

These have been studied for decades, right? And there's been lots of cycles of treatments for these diseases and breakthroughs. And and and so when you get into some diseases that are that have similar, you know, path, you know, similar effects or similar symptoms, but have not been studied that well. The assumption can be on the part of a customer or a client, a drug sponsor, right? A company that is trying to develop drugs to improve that, is that all of the resources that were developed over decades for psoriasis and for atopic dermatitis exist for hydradnata superativa, vitiligo, alopecia, some other indications, other diseases where that amount of Time just has not been put in. And so there's a very poor understanding across the landscape of what you're going to need to do in order to have a successful study in a newer or less well researched indication. Does that make sense?

Yeah, I think that where it is right now, AI there's some things that it does really well that can be applied to to great benefit to this this world, right? But and what those are usually are What are maybe tedious but but simple tasks, right? So counting moles, right? An AI machine that I've seen, you know, they can count six or seven hundred spots on a human body in seconds, right? And so they can detect changes, right? They can detect that there has been a change. That there's been an increase in the number of spots or dots or moles or you know that is valuable. That has incredible value. But the the gap between that, between counting or imaging that and actually diagnosing, or to your point, determining like a a treatment plan or any of that, the gap in between those two things is is vast. Right. And and there are already people who are, you know, who companies and and who have aspirations coming in and thinking like, well, we're we we just can give this to AI to make this determination bore based on morphology, right? So based on what is the appearance of the rash or the lesion that they're seeing. And and it just doesn't work that way. They're the the human in the loop is the the terminology that I've been hearing now. in dermatology particularly is going to be important for for a long time. It's it's gonna be important and not only that, but but you can outsource the you know you can hire AI to you know to take pictures or to you know identify some things that doesn't make you not responsible for what the effects are, right? So if it's a if it is a misdiagnosis or if someone misses you know a a cancer or a you know something Along those lines, that doesn't let the doctor off the hook. The AI tool is not responsible. The doctor is still responsible. So we're not going to be able to remove them from that process for a long time, and depending on the the indication, ever maybe, right? But AI has an incredible role and it and it and a future in not only AI, but image processing with AI, right? Image capture. And that's what we work on, right? We work on. What what what has been called you may have heard, you know, dumb tech, right? We're we're supporting manual processes using technology to automate them or to make it easier for the doctor to conduct these assessments rather than promising to replace the doctor, right? We're not gonna do that and and and we're probably never gonna do that. but certainly this company is not focused on that. What we want to do. Is say what does the doctor have to do right now and how can we help them do that more effectively and and win less time and make the cup patient more comfortable? and that's our that's our focus.

I think that what you know what you can what you will have is an you know incredible you'll you'll have the ability to document things right and to catalog them. So as a quick example, I'll I'll tell you. So so skin tone is a significant there's a significant variation in in the way that a disease looks based on how dark or light the complexion of the person is, right? So the signs of disease are actually more visible or more apparent. And that changes depending on the skin tone. So something that AI could be really good at is with the expert clinical diagnosis of a board-certified dermatologist, you could start to create a kind of atlas of in Fitzpatrick skin type, or if you use the Monk scale, you know, Monk scale skin type. 10, you know, erythema looks like this, severe lichinification looks like this. You could you AI could start to organize that information, but the basis of it is is still gonna have to be an expert in skin disease assessment. so that's sort of where I see it going is building off of the collection of these expert assessments and turning that into something that can be leveraged. you know, to educate as well as to create better diagnostic tools.

well, that's hard. I I don't know. I I don't you know I think If you I I think I really do. I I I don't know if I would say this to my younger Eric because I think I, you know, this is sort of a guiding principle. But you know, if you really are focused on solving problems, if you really are focused on you, you take you take the journey where it leads, right? So when I started out, so we have a product called Lama. Right. And it's a it's a lesion area mapping assistant. It really is unlike anything else, right? There isn't an an equivalent in the market. And we we we ideated this for a very specific use case because there was not a way of doing this well, right? There was a there was a particular process that is used in the assessment of hydratinata supertiva that is just not done well. It can't be done well by a person using standard methods. And so, you know, when I when I when I launched the when I started the company, I had never even, it never occurred to me. I didn't know enough about HS or hydroonata superativa to even contemplate something like that, right? And now we can see that it has, you know, great utility across all kinds of dermatological conditions and we can program it for different things. But this was This was driven by the desire to solve a problem. That was the desire. It wasn't to make money. It wasn't to, you know, any of those things. It literally was, well, well, what would make this better? What would make it easier to do this? and so really it is, I've heard this said before, isn't it coming, you know, this is not original thought to me, but fall in love with a problem you want to solve. Right, fall in love with a problem and then work at solving it, that will drive you, that will take you to, you know, places that you never even considered or never even imagined. And so I think that's where that's what's really exciting to me. And when I look back and think, I'm I am surprised where we are right now and what we're building, what we're working on. and I I just never would have dreamt it.

And but but it really just was an you know it grew out of trying to fix something, right? To trying to make a contribution to the the community. So I think that that that has borne out in my life, not just professionally, but in other aspects of my life. you know, if you have good intentions, you you know, usually the results are good.

Yeah. Yeah, absolutely. It's www.tolokin.com. So it's T H A L O C A N. I know it we pronounce it funny, but you know, there you go. I wanted I wanted to be able to have one word dot com. So I I I chose an odd spelling for the for the the company. so yeah, www.tolokin dot com and we you know we we we have you know

All of our products are listed there. a couple of them, Tavos and and Lama, are pretty unique. the the the Tri-ECOA stuff is, you know, there's many competitors. There's many, many other companies that do that. We we take a little bit of a different approach, but there's many companies that do that. and then all the trainings are listed there as well. And we've got some articles and we've got some, you know, other short videos and things like that. So yeah, anybody can check it out and Obviously my name it's just Eric at Toloken dot com, so it's my first name and the company at Tolokan dot com is my email. So welcome to to ping me if there's anything I can help.

No, thank you so much. I really it was fun. I had a good time, so thanks a lot.