AI-Powered Symptom Checker: Your New Healthcare Companion with Kota Kubo, co-founder and co-CEO of Ubie

[00:00:03] Hey, everyone. Welcome back to the Outcomes Rocket. Saul Marquez here. And today I have

[00:00:07] the privilege of having Kota Kubo with us. He is the co-founder and co-CEO at Yubi.

[00:00:14] Koba graduated from the University of Tokyo Graduate School of Engineering in 2013,

[00:00:20] enrolled at the University of Tokyo. He began researching and developing software and algorithms

[00:00:26] to simulate the relationship between doctors, symptoms and disease names. He has a wide

[00:00:32] experience having worked at M3 Corporation and having worked on a variety of software and web

[00:00:38] tools. I'm excited to really help him cover what they're doing at Yubi here on this podcast today.

[00:00:43] So Kota, thanks for joining me. Kota Kubo Yeah, thanks for having me.

[00:00:46] Steve Adubato Absolutely. And so look,

[00:00:48] let's get to it. Kota, tell us about Yubi. What are you guys up to? And tell us about you. What

[00:00:53] got you into healthcare? Kota Kubo Sure. So Yubi is now providing their AI service

[00:00:57] and for the patients and providers and pharmaceutical companies as well. For the

[00:01:02] patient, we provide an AI-based symptom checker. That symptom checker asks about some sort of the

[00:01:09] symptomatic question like a headache or stomachache and cough. And after that, the AI

[00:01:13] suggests the disease name and also the providers and sometimes a treatment way, especially new

[00:01:19] treatment for the rare disease and the specialty disease. And for the provider side, we're using

[00:01:25] the same AI as to the consumers one. And with that service, we improve their doctor's work

[00:01:32] efficiency by making their clinical record automatically. Because before the medical

[00:01:38] interview, the patients go through our AI symptom checker like an intake home so that the AI

[00:01:44] generates a summary of the patient's condition so that doctors can get to know the patient's

[00:01:49] condition immediately. And also at the same time, they can get a medical record and AI suggests

[00:01:56] a disease to the doctors so that we can get training data from doctors at the same moment.

[00:02:02] And for the pharmaceutical companies within that platform, sometimes the pharmaceutical companies

[00:02:07] get the problem of delivering the information, especially for the specialty and the rare

[00:02:12] disease patients because they are not sometimes misdiagnosed and they cannot reach their

[00:02:18] appropriate providers. So that we actually guide the patient to the appropriate medication

[00:02:25] that they can increase and increase their amount of the prescription appropriately. And for me,

[00:02:30] so basically as you said, I was their software and AI engineer. And at that time, I started

[00:02:37] the research of the disease prediction and why I started this services. Actually, when I was

[00:02:46] a university student, I was poor because my father was a go bankrupt. So that's how, yeah.

[00:02:54] At that time, I was poor but I was not so healthy so that I easily went to see a doctor. But sometimes

[00:03:02] and I felt the pain point of the just the cost of the doctor visit, even though in Japan,

[00:03:08] it is not expensive to see a doctor compared to the other countries. But it's still a big pain for

[00:03:14] me. But I was scared of having the serious disease and they just about with, I went to see a doctor

[00:03:21] very easily. Even if they eat a cough, just a cough or just a cold, just flu. So that, but I realized

[00:03:29] it is sometimes not necessary. And the drugs that doctor gave me was actually the same as the

[00:03:35] OTC drugs. And I realized that at that time also there was a deep learning AI boom. So that's,

[00:03:42] I just come up with the idea to combine the AI to solve that problem, just guide the people to

[00:03:48] the appropriate medical care. With that service, we can save the time of the medical appointment

[00:03:55] and the authority because we can save the doctor's resource as well.

[00:03:58] Ricky No, that's really, thank you for sharing that story. And thanks for sharing more about

[00:04:03] the different indications that that the technology could help with patients, providers,

[00:04:09] pharmaceutical companies. And really, as you think about everybody listening today,

[00:04:14] who would you want to focus on? Who would you say right now is your, while you can help all three,

[00:04:21] who would you say is your main one that you want to be able to help the most with what you could

[00:04:25] offer today? Yes. The most priority of our business is of course the patients and the,

[00:04:33] we're improving the error, disease prediction accuracy. So we are focused on the accuracy,

[00:04:39] especially for the patient with the rare and special disease. Even though there are a lot

[00:04:45] of information in the web, but sometimes for the patient too, it's very difficult to recognize

[00:04:51] their serious disease, rare and speciality disease because it's not typical for them.

[00:04:56] So that's their easily misunderstand. So that's their, it's huge gap. So that's we

[00:05:02] guide them to the appropriate care. Got it. Thank you for that. And there's certainly a lot of

[00:05:08] promise to technologies like these with the use of AI. One thing that keeps coming up, Kota,

[00:05:15] is really, as we get more AI into healthcare, the type of AI being used. So is this NLP? Are you

[00:05:23] using Gen AI? Is it a subset of a different type of AI? Do you care to expand on that?

[00:05:29] Yeah. Great question. So the, basically we use their more like a statistic way of using the

[00:05:35] Bayesian network. And after that, we started to use the reinforcement learning. So it's a more

[00:05:41] like a machine learning way. And recently we combined their generative AI technology as well.

[00:05:47] So that's, it's a very mixed one. One of our competitive advantage is getting the real

[00:05:53] clinical data since we have the huge platform of their provider size. We have the partner with

[00:06:00] more than 1700 Japanese medical organizations and they gave us a lot of data including the

[00:06:08] diagnosis. So that's where we can refine our AI with that data. Thank you for that, Kota. Folks,

[00:06:14] there's huge promise in what UB is doing for all of the stakeholders that Kota mentioned. I know

[00:06:20] we're almost out of time here, Kota. Today was just like a nice preview to your business, to you

[00:06:25] and the great things that you're up to. What call to action would you leave our listeners with? And

[00:06:30] where can our listeners and viewers learn more? The U.S. business from the end of the 2022,

[00:06:37] and at that time we launched our service for the patient and the user base is now 2 million.

[00:06:44] And with that platform, we are providing the promotion service with the pharmaceutical

[00:06:50] companies. So what differentiates us from the other promotion world is we are just providing

[00:06:58] of course, the patient access, especially for the speciality disease and rare disease. But at the

[00:07:03] same time, we provide evidence of whether the encouraged patients go to see a doctor or not.

[00:07:12] After our validation, the pharmaceutical marketers can get to know the eye of our product. So that's

[00:07:19] very reasonable for the pharmaceutical marketers compared to the old media, just an impression of

[00:07:26] quick word. So it's totally different from the service. So that's the pharmaceutical marketers

[00:07:33] who is very interested in our product, please reach out to us. That's great for all the pharma

[00:07:39] marketers out there looking for a platform to help get access and more success. Reach out to

[00:07:46] Kota and the team at UB. Kota, such a pleasure to have you here on the podcast today. Thank you very

[00:07:51] much again.