Kyoto Startup Digest 2018 – HacarusKyoto Startup Digest 2018 – Hacarus

Kyoto Startup Digest 2018 is documenting the Kyoto startup scene through a series of articles and videos by the following members of the KYOTO Design Lab and interns from the University of Texas.

Extending human life with Artificial Intelligence

In 2014, former Sony Playstation engineer and three-time startup founder, Kenshin Fujiwara began his 4th startup in Kyoto. Hacarus, a company with a unique Artificial Intelligence technology and big goals, set its sights on helping people live long, healthy lives.

Hacarus started out as an IoT company with a hardware product, smart kitchen scale that would collect and store food nutrition data for diet optimization. After not reaching their goal in a crowdfunding campaign, the company pivoted and decided to focus on the AI technology behind the companion app. The sparse-modeling data analysis AI platform being developed was directed towards providing solutions in medical equipments, medical research, manufacturing optimization, and overall defect detection.

Data Scientist Keiichiro Takahashi

Hacarus’s core service is Hacarus-X, a non-deep learning AI that is able to collect data, analyze data, and provide predictions and results that is easily understandable through sparse-modeling. This differentiates Hacarus-X as its algorithm does not require big data. Instead, it collects small amounts of data and extracts the most meaningful and essential parts for analysis. This is not only effective but convenient as in many cases, collecting large amounts of data is too difficult, too costly, or simply not possible.

Health Fitness Programmer Tomomi Kikumoto

Hacarus Fit Platform is a toolkit for medical, health, and fitness industries that is able to work as a standalone system or with existing systems. The platform includes an application programming interface and a software development kit that connects with Hacarus-X and makes developing an AI service easy. So far, the platform has been used by companies wanting to offer personal health record services or add healthcare features to services already being offered to existing members.

As opposed to companies operating in the field of big data, Hacarus keeps its edge by focusing on sparse modeling to save time and money for their customers. Hacarus acquired seed funding of approximately half a million USD in 2016 from 7 angel investors and VCs. The company followed that up with another half a million in July 2017 in what Kenshin calls a “pre-series A.” As Hacarus continues to move forward in AI data analysis, the company anticipates successful strides in piecing together the puzzle of extending human life.


Interview
In order to learn more about Hacarus’s unique AI technology, we sat down with CEO Kenshin Fujiwara along with Harvey Cornell, summer intern from the University of Texas at Austin.

What was the initial idea when you started Hacarus?

Kenshin: We have to do some kind of exit. We have five institutional VCs and couple angels. But my ultimate goal is that I want to extend human life beyond 100 years. If we want to live on Mars, 100 years isn’t gonna be enough. You need to reach 200 or 300 if you want to live outside this solar system. So first, I need to extend human life, then I can work on something else. I’m solving one core problem and to do that, I need a lot of people and lots of resources.

What prompted Hacarus to move from IoT to SaaS and AI technology?

Kenshin: Making hardware is really difficult. That’s why we had to make the choice to discard the hardware part. That was the first pivot we had to go through. Then we were selling this companion smart phone app to enterprise users, like fitness companies, healthcare companies, or big enterprises that have 1000+ employees, but that didn’t go well either. We have one existing customer from the fitness industry called Tipness who is one of the top 5 fitness companies here in Japan. But we started focusing on the AI, which was quite interesting because we’re not using deep-learning technology. Because our algorithm is so unique, we decided to use this technology and build a service around it.

What is special about Hacarus’s technology?

Harvey: When you think of deep-learning, the natural way to think about it is to collect thousands of samples of data and slowly refine the algorithm based on all these test cases. The unique part of our algorithm is that it’s not based on taking in these meta-data sets. Instead, you take in a relatively small number of data sets and through that, the algorithm is able to predict or estimate what the result would be for a new outcome. What we use is called sparse-modeling, that’s the estimation or predictive aspect that’s involved with smaller sets of data. So instead of using big data, the massive amounts of data being used by many companies like Amazon, Google, all these major corporations or even a lot of other AI tech startups, Hacarus is using small data giving us a really big edge.

What challenges have you faced? Anything particularly related to Kyoto?

Kenshin: First is hiring. We aren’t in Tokyo or San Francisco, so in our case, it’s very hard to get talented engineers, especially data scientists. This is one reason why we started hiring in the Philippines. As for fundraising, I dont think it’s difficult anymore because we have too many investors and too few startups so the power balance has changed in the past 10 years. Another thing is regulations. Technically, AI cannot diagnose patients. It can only provide information and only doctors can diagnose patients according to regulations. We have to work with the government, regulations, and guidelines. That’s the process we have to follow especially in the healthcare or medical domain.

Do you have any plans for scaling up?

Kenshin: We have to do some kind of exit. We have five institutional VCs and couple angels. But my ultimate goal is that I want to extend human life beyond 100 years. If we want to live on Mars, 100 years isn’t gonna be enough. You need to reach 200 or 300 if you want to live outside this solar system. So first, I need to extend human life, then I can work on something else. I’m solving one core problem and to do that, I need a lot of people and lots of resources.

Why Kyoto?

Kenshin: I did my previous 2 startups in Tokyo and sold both of them there. If you do a startup in Tokyo or San Francisco, you don’t have to think, you just do it faster than everyone. Raise tons of money and be the first one to put the flag in the domain. That’s the strategy to win the game. I didn’t like that idea and got tired of it so I came back here and decided to do something different. Kyoto is known for being a hometown to life science companies, especially around Kyoto University. There are many biotech startups happening here so I knew life sciences and healthcare is the core strength of this city. Second is manufacturing. We have Kyocera, Omron, etc, all these big manufacturing companies. I wanted to do business around these 2 sectors. Also for the sparse modeling technology, the professor we are working with is named Mr. Masayuki Ohzeki. He used to be a professor at Kyoto University, and that’s why we got an investment from Miyako Capital.

Do you have any advice for future founders in organizing their team?

Kenshin: Just do it without thinking. Of course if you want to raise money from VCs or even angels, you need to have some sort of business plan to talk about. But that’s only the paper, it’s all in your head. It’s just assumptions, and things will change as you work on your business. Learn from your failures but don’t make big failures. Make small mistakes and learn from them. That’s what I call, making small pivots everyday, not a big pivot or big failures.

Do you have any advice for startups and entrepreneurs before they venture into the Kyoto area?

Kenshin: Find something you can feel passionate about, otherwise you’ll quit easily. Find something that interests you so you can keep working on it for a long time. It doesn’t matter if you’re coming to Kyoto, Tokyo, San Francisco, or even Austin, that’s the most important thing. Don’t do a startup for money. I made that mistake for 15 years. I did my first and second startups just for money and kinda wasted my life for 15 years. This time it’s for my own interest.

Check out Hacarus here


KSD 2018

00|Introduction
01|Nota Inc.
02|Makers Boot Camp
03|Hacarus
04|Kyoto Startup Summer School

Related Links
Related Posts
Share