Disrupting Japan

Sometimes it seems like Japan just can't figure out medtech startups. In theory, Japan should be one of the most active startup markets for medical technology. In practice, however, there are relatively few success cases. Today we are going to meet one of those success cases and talk with Sho Okiyama, MD and founder of Aillis, who has developed a minimally invasive flu and covid test that is already in widespread use in Japan. Hopefully, no one will ever need to stick another swab up your nose. We also talk about Sho's time as a rural doctor, what's straining Japan's healthcare system, and how to fix it before it breaks. It's a great conversation, and I think you'll enjoy it. Show Notes Why influenza diagnosis has been so resistant to AI You’ll never get another swab stuck up your nose What practicing as a rural doctor taught Sho about Japan's healthcare system Some Japanese still turn to Shamanism for medical advice Starting an AI startup without a technical founder Building a medical training dataset from scratch The challenge Anthropic and OpenAI face in medical tech Aillis's plans for global expansion Why there are so few medtech startup in Japan and how to fix that How to drive meaningful change in Japanese society Links from the Founder Everything you ever wanted to know about Aillis Learn more about Evidence Finder Connect with Sho on LinkedIn Friend him on Facebook Follow him on Twitter @ShoOkiyama Watch Sho's TEDx talk on AI development in modern medicine Leave a comment Transcript Welcome to Disrupting Japan, Straight talk from Japan's most innovative founders and VCs. I'm Tim Romero, and thanks for joining me. Japan has a real problem with life science startups. I mean, there's no shortage of brilliant ideas and patents. There's no shortage of support programs and investors looking to invest in Japanese medtech and biotech startups. Everyone in Japan wants this to happen. Everyone knows this is about to happen. And yet, everyone is also a bit frustrated at just how slow this is happening. But of course, it is happening. So, today we sit down with Sho Okiyama, medical doctor and founder of Aillis, who has developed an AI-powered throat camera that can actually diagnose both influenza and COVID. It's the first device of its kind ever approved as a diagnostic medical device, and it is already in widespread use in clinics and hospitals in every prefecture in Japan. It's really an amazing success story that is just getting started. Sho and I talk about what it really takes to bring a new medical device to market in Japan, what it's like to be the only doctor on a string of remote Japanese islands, and how we're going to get more Japanese doctors to start startups. But you know, Sho tells that story much better than I can. So, let's get right to the interview. Interview Tim: So, we're sitting here with Sho Okiyama, the founder and CEO of Aillis, who is using AI to make medical diagnosis faster and more accurate. So, thanks for sitting down with me. Sho: Thank you for having me, Tim. Tim: I'm really looking forward to this, because I think there's so much innovation going on in biotech and medicine in Japan right now, and what you guys are doing is really interesting. So, you've got two products, and I'd like to kind of cover them briefly. First, you have the Nodoka diagnostic camera. Sho: We produce Nodoka, which is a medical device. This is a pharyngoscope. In other words, it's a throat camera. So, whenever you go to the doctor, the doctor would take a look at your throat, right? You know, find out whether it's reddish, swelling, or something like that. But all those inspections were based on knowledge inside the doctor's brain. So, what we developed was the database of the pharyngeal photo. Tim: So, when a doctor is looking at someone's throat, they're looking for all kinds of different things. But Nodoka, as I understand it, is looking for two very specific things, right? It's able to diagnose influenza and recently COVID as well, right? Sho: Yes. We used to diagnose influenza and COVID using PCR or similar methods. Tim: How accurate is it compared to the PCR test? Because I would be very, very happy if I never have to get another swab up my nose for testing these things. Sho: Yeah. Nodoka has gone through a clinical trial and has proven that it has approximately equivalent accuracy compared to the antigen testing. Tim: That's fantastic. And it's in widespread use already, right? Sho: Yes. It has been introduced to more than 2,000 hospitals or clinics in Japan. Tim: And is the driving factor just patient comfort? Nobody likes getting the swab up the nose? Or is it more accurate? Is it faster? Sho: Yeah, it's faster. And of course, it's less invasive, right? And what's more, Nodoka only has indications for two different diseases as of now, but it has the potential to have more indications, like cancers such as tongue cancer or pharyngeal cancer. And we also wrote a paper that the AI is able to detect signs of diabetes only through the image of the pharynx. Tim: Really? Okay, I want to get back to Nodoka because it's super interesting. But before we do, I want to talk just really briefly, introduce your other product, which is Evidence Finder. Sho: So, this is an LLM-based tool we developed for doctors. There are tons of medical information coming out. There are hundreds and thousands of papers published every day. So, doctors need to keep up, right? But it's a huge burden for doctors. So, what Evidence Finder does, it's a chat-based tool, such as, ChatGPT or Gemini, but it cites scientific papers. We use multiple techniques to prevent hallucination, which is very, very important in the field of medicine. And it also gives the doctors concrete evidence why the AI had said what he has said. Tim: So, is this more of a research tool, or is it something that could potentially be a diagnostic tool? Sho: It could be used in both ways. If the doctor were to use it in clinical settings, it needs to be approved as a medical device, which also could be the case for us developing the medical device version of Evidence Finder in the future. Tim: Well, I understand that getting certified as a medical device is very much on your roadmap at the moment, right? Sho: Yes. Tim: Okay, I really want to dig into that. But before we do, I want to back up a little bit and talk about you. Sho: Thank you. Tim: So, you graduated from University of Tokyo Med School in 2010. And after graduation, you ended up practicing in some very rural areas of Japan, right? Sho: Yes. Tim: I mean, University of Tokyo Medical School is the most prestigious medical school in Japan. Graduates go into top-tier internships and hospitals all over Japan. Why did you decide to go to Okinawa and these rural areas and work as a rural country doctor? Sho: Yeah, I started my residency in Red Cross Hospital in Shibuya. This is a huge hospital. It has like 20 doctors on duty at midnight. And you can perform MRIs, CAT scans 24/7. But the hospital was too developed for myself. And if I were to work there for 10 years, I would end up as a doctor who is only able to work a developed hospital like the Red Cross Hospital in Shibuya. So, I found in myself that I needed to go to rural areas or small clinics to develop my physical examination skills, as well as not ending up as a niche specialist. I wanted to be a generalist. Tim: That makes a lot of sense. So, I mean, I could see that on one hand, going into prestigious hospitals, it's a very smooth escalator. On the other hand, it sounds kind of like the medical equivalent of being a sour man. Sho: Oh, yeah, you could say it in that way. Tim: I mean, things are decided for you. You sort of get pushed into doing one specific track or rotated through. So, how long were you in Shibuya before you made that call? Sho: I was in Shibuya for three years. And then I went to Ishigaki and Hatsuka Islands for one year. And Hatsuka has only 500 inhabitants on the island. And there is, of course, I'm the only doctor on the island. So, you have to be on call for 24/7. If there are emergency patients, I had to call the helicopter and ride the helicopter with the patient to the main island. Tim: Right, right. Because I imagine you're responsible for a number of the surrounding islands. Sho: Yes, yes, we are. Tim: The lack of rural doctors in Japan, it's a real problem now. Do you see technology as being able to kind of address this? For example, the Ministry of Health and Welfare is pushing hard on things like telemedicine and technology sandboxes to allow new technology to be experimented with. Do you think technology is the solution or is it more getting human doctors interested in going? Sho: It's both technology problem and the doctors themselves. But it's less difficult to solve this problem in a technological matter because human beings or doctors don't switch their ways of thinking instantly. But technology, it's changing every day, every week. Tim: Yeah, that's true. You'll certainly make faster progress on the technology front than on the human front. Sho: Yeah, but the decision I made to go to the island, this was 2012 or so, 15 years ago. So, there was less technology, there was no telemedicine at all. So, I had to go there, I went there, and noticed that there's a huge gap between the medicine practice in Tokyo and on the island. But I was taught in the university that Japan is the top one country which has the least discrepancy between the medical level within the nation. Tim: Really? I believe that, I could see that. That's something to be proud of. Sho: But the reality was that there still is a huge gap. So, after going to the island, I want to lessen this discrepancy. And medicine is not for the rich people. It's for the humanity itself. ...

Last month I gave lecture at Globis University on what it takes to build an AI startup today. It's no longer early days for AI, and most founders don't have the connections and resources that drive toady's multi-billion dollar seed rounds. However, as I detail, they still have several paths to success. After the lecture I am joined on stage for a panel discussion by Reiji Yamanaka, the managing director of the Kibo Impact Investment Fund, and Kelvin Song, the program director of the Globis MBA program. It's a fascinating discussion, and I think you'll enjoy it. Leave a comment Transcript Welcome to Disrupting Japan, straight talk from Japan's most innovative founders and VCs. I'm Tim Romero, and thanks for joining me. I have a special in-between episode for you today. A few weeks back at Globis University, I gave a lecture to aspiring founders on the best way to start a generative AI startup right now in this time of intense AI competition and funding levels. I cover the different AI business models, promising application spaces, and how to know if you've got an AI startup idea with a good chance of success. Now, the first 30 minutes of this episode is the lecture itself, and then I'm joined on stage by Reiji Yamanaka, the managing director of the Kibo Impact Investment Fund, and Kelvin Song, the program director of the Globis MBA program. And we dive even deeper into these ideas and also talk about how generative AI is likely to affect us all. I hope you enjoy it. So let's get right to the presentation. Presentation Today we're going to talk about how to build a generative AI startup and some important things to keep in mind if you actually decide to do that. Now, before I tell you what we're going to cover, I want to kind of tell you what we are explicitly not going to cover. So first, we're not going to talk about the transformative nature of AI in general, the explosive growth of the market. There's already way too much chatter about that, and I assume if you're even thinking about starting an AI startup, you already know it. Second, I'm not going to offer general advice about starting and growing a startups, although this is a topic that's very close to my heart. I want to focus on what can add the most value to you in this particular seminar. If you want to talk about general start advice, talk to me later. I'll point you in the right direction or ask questions afterwards or during the panel discussion. We'll begin today by talking about four common exit and growth strategies. This is a bit unusual. I don't normally recommend that seed or pre-seed companies focus too much on exit strategy, but these are not normal times. With generative AI, you need to plan your end game from the very beginning. We'll spend the bulk of our time talking about actually building your AI startup. We'll cover some key strategic considerations, and also talk about a few of the most promising targets for AI disruption. Does that sound good? Well, before we get to it, why should you listen to me? And that's a totally reasonable question. So, I've been in Japan for, wow, over 30 years now. Currently a partner at Jira Ventures. It's $300 million corporate venture capital firm that invests in green energy, next generation energy, generation technologies. But in my time here, I've started four of my own startups I've sold two, bankrupted two. So, 50 50, not too bad as far as startups go. I've done a lot of angel investing. I've taught entrepreneurship and corporate innovation at New York University's, Tokyo Campus. I've brought foreign startups into Japan as a country manager. I was tapped by TEPCO to come in and help them spin up TEPCO Ventures. I left TEPCO to run Google for Startups, Japan, swearing I would never go back to energy CVC. After four years at Google, I decided to go back to energy CVC because right now what's happening in energy is just fantastically exciting. Oh, and I also run a podcast called Disrupting Japan, where we sit down and we talk with Japanese startup founders and VCs, not so much about their specific company or portfolio, but what it's like to be an innovator in a culture that really prizes, conformity and what problems these startups are trying to solve. So, you know, please like, and subscribe and all that. Okay, let's get into it. Let's talk about your exit strategy and the possible business models that stem from it. Now, as I mentioned before, this is not the usual way of doing things. At seed stage, founders don't normally need to think too deeply about the exit strategy. If you build a rapidly scaling business that adds value to your customers, M&As IPOs, both paths remain open to you and it'll eventually become clear to you which is your best option. It's not something you normally try to optimize for, not at seed anyway. However, these are not normal times, and if you're building a generative AI startup, your exit strategy will greatly influence how you run your business. So, the four exit and growth models in increasing order of complexity are the Peter Pan, riding the hype, rapid customer growth and sustainable revenue growth. Now, we're going to move really quickly through the Peter Pan and riding the hype, since they don't really require growth at all. And honestly, only sustainable revenue growth needs a path to profitability. So we'll be spending by far the most time on that one because sustainable revenue growth, it's the most flexible, it's the most complex, and in most cases, it'll give you the greatest chance for success. Now, as a founder, you need to understand your options and consciously make a choice of what your goal is. And on the other hand, if you're thinking of joining a startup, you need to make sure that you're a hundred percent aligned with the strategy that startups actually executing. Not just the one they're saying they're executing. Now, before we get into the details, I need to point out that, that even if you fully embrace the potential of AI, and it's transformative nature, the VC market has never seen anything like this level of hype and focus. We're already way beyond what we saw during the dotcom area. So, just how hot is the AI market right now? Well, it's stupidly hot right now. Between 60 and 70% of all VC investment is flowing into AI startups. In fact, if you remove the AI investments, VC investment is actually down sharply. So yeah, these are not normal times. So, let's talk about exit and growth strategies. First, the Peter Pan, also known as I'll Never Grow Up. Now this is not a serious strategy, but sadly, it's the one that most startups are executing either consciously or unconsciously. They raise some pre-seed, maybe some seed money. They play with some cool AI tools, they ship a me too product, and then they disappear in a year or two when they can't raise a Series A. Now I include this for two reasons. One, because like a surprising number of founders seem to be fine doing this. Some people just seem to want to have a founder experience and run what I've heard called a starter startup, which I mean, if that's what you want to do, fine. But it seems like kind of a waste of everyone's time. We are here at the beginning of what might be one of the most important transformative shifts in technology. And man, it makes sense to swing for the fences. Don't waste your time chasing small dreams on this. And two, I include the Peter Pan model for those that are thinking of joining an AI startup. If you're thinking of joining a team that is not clearly executing one of these other strategies, they're probably never going to grow up and you're probably wasting your time with them. Okay, let's move on to more profitable strategies. Our second strategy is to ride the hype. Now, I mean, all startups need to do this to some extent. We see hype driven business plans in every hot startup sector, but I mean, AI has everything dialed up to 11. Riding the hype is a pure play. This is pure marketing. Your product is your stock. It's a very good play for the right kind of team. So, what do I mean? We saw safe super intelligence raise a pre-seed round at a $5 billion valuation. They got a nice step up to a $30 billion valuation at seed. Thinking machines just raised 12 billion. Open AI acquired Johnny Ivey’s IO for 6.5 billion on the promise that they would eventually build something cool. Now, these are obviously really big numbers, but if you haven't worked in VC or startups for a while, it might not be apparent just how bonkers these numbers really are. So, when investors invest in seed startups, they're generally looking for a path to a hundred times return. They're not expecting a hundred times return, but they need a path to it. Now, even assuming these are somehow special, they're only looking for 50 times return or 30 times return. This is implying a trillion dollar exit, which would make it not only the biggest startup exit in all history, not only twice as big as the biggest startup exit in all history, but twice as big as the next five biggest startup exits in history combined. And that's not, keep in mind, this is not a bet on the technology. This is a bet on one particular company. So, yeah, thing things are things are a little frothy. Now, these valuations are based on the star power of the founders, and that's important to note that none of these companies have an actual product or even a prototype yet. So, shipping a product is actually a very dangerous step for hype driven companies because that's when people are able to compare the hype and the reality, and very often the reality just doesn't live up to the hype. But anyway, if you are extremely well known in the industry and you can line up an A-list VC to give you that credibility with other investors and to help you exit when the time comes, this could work for you, ride that hype as far as it'll take you. Generally however,...

Last month I spoke on a panel about the future of climate tech. I was joined by Emi Naganuma, the founder and General Partner of Apprecia Capital and Richard Youngman, the CEO Cleantech Group, with Michael Matsumura of Scrum Ventures moderating. Right now is both a challenging and an exciting time for climate tech innovatoin. It's a fascinating discussion, and I think you'll enjoy it.

Progress is not only slower in Japan, it is often different. Looking at the numbers, it's clear that venture capital is even more male-dominated in Japan than it is in the West. Our guest today explains not only how that's changing, but how she's changing it. Sophie Meralli is a Partner at Shizen Capital and co-founder of Tokyo Women in VC. We sit down and dive deep into the keys to developing a creative, global mindset among Japanese founders and VCs, the role immigrants have to play in developing Japan's startup culture, and what really works in changing, not only minds, but actions related to the role of women in startups and venture capital. It's a great conversation, and I think you'll enjoy it.

Fifteen years ago, University-run venture funds were all illegal here in Japan, but today a higher percentage of major Japanese universities have VC funds than in the US or Europe. Today we sit down with Kei Furukawa, a partner at the University of Tokyo IPC, a $300M venture fund, and we talk about the unique role these funds play in Japan, how they drive innovation in rural areas, and why he has to talk professors out of becoming startup CEOs. It's a great conversation, and I think you'll enjoy it.

Today we are going to break down some startup stereotypes. I sit down with Kunio Hara, co-founder and CEO of Beatrust and break apart the stereotypes of the uncreative Japanese enterprise and the young startup founder, and Kunio explains how Beatrust is already teaching old dogs new tricks. It's a great conversation, and I think you'll enjoy it.