English subtitles. Eight Values interviews Masashi Yokota, the CTO of RESTAR. We cover creating an international engineering team in Japan and what it means to be a CTO. We ask about why he chose Elixir as his backend programming language and why it helped him hire some incredible engineers.
Links
Topics
- (00:00) - Intro
- (00:16) - University AI Research
- (04:20) - CTO's ideal engineering team
- (08:00) - Getting first clients
- (10:43) - Motivation to continue in early stage
- (12:01) - Learning to work with co-founder
- (13:12) - How they grew the business?
- (19:00) - Why use Elixir for programming?
- (21:51) - The Role of CTO
- (25:01) - How to manage people
- (34:40) - Advice for CEOs and CTOs
Transcript
Ryohei Watanabe: So I'm here with Masashi Yokota, the CTO of RESTAR. Thank you, Yokota-san. I'm looking forward to working with you today. Thank you. Before we start the RESTAR, I want to talk about something. What were you doing before RESTAR?
Masashi Yokota: Before RESTAR, I was a graduate student at Tokyo University. I was doing AI research, machine learning. My specialty is VQA, which is multi-modal learning. For example, image recognition, image processing, and natural language processing. I was doing research that was more human-like and could be understood.
Ryohei Watanabe: Why did you research AI?
Masashi Yokota: It's a difficult question. Simply put, it was a boom. I was doing AI research, and I just wanted to say it. But when I tried it, I was not a computer scientist when I was in college. I was originally studying hardware. But as I studied hardware, I wanted to do more programming, and I wanted to study computer science. So when I went to college, I was in a different direction.
Ryohei Watanabe: At that time, when you were thinking about what to study and you chose AI. Did you have any career goals in mind or anything you wanted to achieve?
Masashi Yokota: That's a good question. At first, I really wanted to be a code craftsman. I didn't want to do anything other than coding. That was my first plan. I didn't care about anything else. I could have been a researcher, or a web engineer, or a machine learning engineer, as long as I was coding. That was my first career goal.
Ryohei Watanabe: What do you think about that career image compared to now?
Masashi Yokota: In that sense, I think it's completely different now from when I first joined RESTAR. When I joined RESTAR, I was the first engineer. I had to make the product. It was really fun at that time. I was constantly coding every hour I was awake. It was really fun. It's still fun now, but I can't really do coding now. I think I have to be in the CTO position. I have to do people management, project management, and other things. I'm also in a position of taking on the role of the CTO. I think I have to think about issues with the company as a whole. In that sense, I think my main job is to build more team building.
Ryohei Watanabe: You said that CTO is the main team building. What kind of team do you want to build?
Masashi Yokota: I think I'm still not able to achieve what I'm aiming for. I think each team should be professional. I think it's good to have a team that respects each other and makes use of their strengths. I have a lot of shortcomings. I'm not good at fine checks, I'm not good at schedule management. I'm also a project manager. But the current team members are supporting me, so I'm able to develop the product. I think it's best to have a professional engineers, where each person is good at something. A strength like, good at security, who knows the language of Elixir, or who knows a lot of things. I think it's best to create an environment where engineers can make the best use of each of their strengths to help other members. I think it's best to create an environment where engineers can do what only they can do. Of course, it's important to improve your weakness, but I think it's more important to make use of your strengths.
Ryohei Watanabe: If want to build a team that can take advantage of their strengths, you have to think about this goal while you choose your team in the interview stage and in recruiting. Is that difficult?
Masashi Yokota: That's right. It's quite difficult. I want each engineer to be a professional. If you don't have this, you don't have your strengths. In the interview process, you need to have basic knowledge. For example, you need to check how much you understand the technical words you use in the interview process. If you are a professional, you should be responsible for your basic knowledge and what you say. You should understand it and say it. I'm looking at whether or not you can communicate properly with the team. I think that in a sense, we tend to ask questions in detail. But I think that's really the team's color.
Ryohei Watanabe: I think that's one of the professional factors we think about. I'm quite concerned about that. I'd like to talk about the team and the CTO a little later. I'd like to talk about the first stage of the list.Did you accept the company, the customers, the list at first? Did you understand them?
Masashi Yokota: In that sense, in my impression, the concept was praised at the very beginning. But it was a pretty old company. They didn't want to be our first customer. So it might be a good service, but once someone else buys it, I'd think about it again. I was very irritated. I wanted them to decide for themselves whether to buy it or not. On the contrary, in this phase, there are many customers who have already bought it. We have a good reputation in the industry. We can get customers based on our demos so I think it's a great phrase. Compared to the beginning, I think it's easier to do business.
Ryohei Watanabe: How did you convince the first customer?
Masashi Yokota: I didn't convince them. I think it was the networing power of the CEO, Uno-san which was very big. The first customer was the senior of Uno-san's high school. The company was motivated to buy it because the senior worked hard to convince them. I see. There was a small customer before that, but I think it was the first time a big company bought our service properly.
Ryohei Watanabe: When the first customer was not there, what was the motivation to continue?
Masashi Yokota: The motivation was just that it was fun. The office at that time was smaller than the room I'm shooting now. If you stretch your hands, you can easily touch Uno-san. It was like a single room. I've been working in such a room. But it was really fun. I thought Uno-san's business idea was very good. In a sense, there was no limit. I talked to Uno-san straightly about what I wanted to do. I was just excited to make a big business.
Ryohei Watanabe: It was just fun. Did you feel comfortable working with Uno-san from the beginning?
Masashi Yokota: It was very easy to talk to, but there were also many conflicts. We were both from different industries. I had never worked in a company before, but I was an engineer. Uno-san was a company that was a little different from the ones I worked at. I understood what Uno-san was thinking. Uno-san understood what I was thinking and how I could work.
Ryohei Watanabe: But until we understood each other, the way we worked became stressful. But we talked a lot and gradually corrected it. After you and Uno-san collected the first 10 companies, was it difficult to grow? Where did you start? There were 10 companies at the beginning. How did you increase them after that?
Masashi Yokota: The first branch was Wada, then Kobayashi and Nakashima. I think the sales team was completed. I think it changed a lot from there. It takes a lot of time to talk to the company. It takes a lot of time to approve the company. It takes at least three months to approve. We needed a lot of people. Uno-san is not a salesman, so we didn't know what to do. I think it was a good time for the new employees to support us. I think it was a good time for them to start. I think it was going well from there. I think it was a good time for them to start. I think the best team members joined us. I like the team members who are proud of their friends every day. I respect them as a member. For example, engineers don't think about coding. They don't think about business. But for the listeners, they think about business. They think about the business plan. I'm not a professional, but as a member, I'm sure they'll tell me if I have any concerns. I think that's why the team members were able to get ideas. I think we were able to hire good members. I don't think Uno-san and I could have done it without Remetis.
Ryohei Watanabe: How did you hire such good members?
Masashi Yokota: There are various paths. As a senior engineer, John and Adam, who are working as a senior engineer, They applied directly from our career website form.
Ryohei Watanabe: Why did they apply?
Masashi Yokota: They applied for Backend in Elixir. There weren't many Japanese companies developing in Elixir. They applied by chance. I think it's a little different. That's the biggest thing. They have a lot of things that I can't do. They are engineers who can be trusted. They are really reliable as businessmen. They applied for Backend and I think they're really lucky. I think this is luck. Other members are... There is a job description site for foreigners who want to work in a Japanese company. Some people have published a job description and applied. Other people introduce themselves through agents. Everyone is very good. They are professionals. Like other members, they have a high level of respect for the listeners.
Ryohei Watanabe: I really enjoy every day. I think it's a common impression. For example, if you don't make Backend with Ruby, Java, and Python, It's hard to hire a good engineer. In the case of RESTAR, they decided to choose Elixir. I think there was an influence that John and Adam came. Do you think other companies should use Elixir? Or the most popular language?
Masashi Yokota: I think this is a strategy. Of course, there are pros and cons. The reason I chose Elixir is... Originally, I was developing Ruby. I wanted to return a lot of data in a faster response. I was looking for a language faster than Ruby. I also wanted a language that would be easy to maintain. My friend has been using Elixir since it was released. I think it was released about 5 years ago. I think it was about 6 or 7 years ago. My friend has been using it for about 2 years. I've always heard good reviews. So I wanted to try Elixir. And of course, I wanted to try Golang. After doing a lot of prototypes, Elixir was the best. Another thing is that Ruby engineers are often in Japan. I think it's standard to develop Ruby. However, if you hire Ruby engineers, you can compare a better Megaventure and a RESTAR. Then you don't have the motivation to choose a RESTAR. Because it's a really small team. I'm the third engineer with only me and Uno-san. At first, the salary is not that high. Then why do you enter? I think it's better to be technically interesting. In a sense, RESTAR's business is a niche business. I thought that it would be a motivation for a better engineer to be a little niche in the tech stack.
Ryohei Watanabe: Do you think that hypothesis was correct?
Masashi Yokota: I think it was correct. The fact that Adam and Jan applied directly is the proof. Very good engineers came.
Ryohei Watanabe: Next, I would like to talk about the role of CTO, Yokota-san. What is a good CTO? What kind of CTO do you want to be?
Masashi Yokota: I think the position of CTO is very difficult. The reason is that if the phase is different, the ability to be desired will change. In other words, the ability to start a business is important. After the series, team building and team building may be necessary. After the CTO, you have to explain to the owner. I think that's the ability. If the phase changes, the necessary knowledge and ability will change. I think that each company is different in what is important at this stage. If I tell you what kind of CTO I want to be, As I said earlier, the team I'm aiming for is a professional team that everyone can use their strengths to help each other. In order to do that, I think it is necessary for everyone to work together. I think it's good for everyone to be able to pull together. In that sense, I want to be able to say what I want to say in a flat way. In addition, it is important that everyone can work comfortably. I think that people management is also a point as a CTO. I've been studying people management and failing it. But people management is also a phase that is left to others and focused on technology. I think that's different depending on the company. But now, I'm at a stage where the team is getting bigger. Even if I adopted a manager, I'm the one who feeds the manager. I'm studying people management because I want to study it.
Ryohei Watanabe: How did you become a good people manager?
Masashi Yokota: In that sense, I haven't become a good people manager yet. The difficult thing about people management is that you have to understand that person. In a sense, I think that you shouldn't communicate with everyone in the same way. Our team is international and we have a lot of experience from various countries. For example, even if you get one feedback, if you make a mistake in how to convey it, you will feel more stress than necessary. I'm looking into the culture of that person and the communication style of that country. I'm looking into the day off. I understand the personality of that person and make a little customization. But I don't think it's good to be kind to this person and strict to this person. In a sense, I think it's important to be kind to this person as a technique. People management is always important to communicate what you think and what the team wants without miscommunication. In that sense, I think it's important to understand that person.
Ryohei Watanabe: I think that the team of RESTAR is increasing, especially the team of software engineers. Yoko-san is Japanese and the team is made up of Americans, Filipinos, French, and British. I think it's a pretty international team. How do you communicate more efficiently?
Masashi Yokota: In that sense, I think it's more important to take it often than efficiently. Our team is basically living in Japan. But it doesn't matter where you work in Japan. For example, there are people who live in Fukuoka, and even sales teams don't live in Tokyo. For example, Hyogo, they live in Kansai. So basically, we work remotely. As a result, the chance to talk gradually decreases. For example, if you communicate in text, it's a little bad. Also, you can't communicate in real time. You don't say something so strict, but you feel stressed. That kind of miscommunication increases. So I think it's important to increase the frequency of communication in the high-level. I think the content to talk about is the next step. So I think it's important to increase the frequency of communication as much as possible. If you reply to comments on GitHub, you can stop commenting there and talk in high-level. Also, I do a coffee break every week to have a casual chat. I don't do anything at work. I just talk for 30 minutes while drinking coffee. I think it's important to have a casual chat and work. I think everyone needs to understand each other. So I think coffee breaks are important.
Ryohei Watanabe: What kind of topics do you talk about at the coffee break?
Masashi Yokota: We talk about various topics. For example, I recently got a girlfriend.
Ryohei Watanabe: You talk about private things!
Masashi Yokota: Or on the contrary, I can't get a girlfriend at all. I talk about where I met her. For example, when I had a day off, I talked about what I was doing on my day off. I don't talk about work that much.
Ryohei Watanabe: What are the advantages of being a restart SE?
Masashi Yokota: I think there are three points. The first is that it is international. I'm talking in Japanese now. My team is international. Everyone can speak English except me. I'm from various countries. Of course, there are members who are not good at Japanese. So it's easy to enter without speaking Japanese. I don't think there are many people who can work in an international environment even though they live in Japan. That's the first point. The second point is that each team I'm aiming for has a professional member. Even if you are a junior member, you only have the potential to be a good engineer. There are only excellent members. Our team's code review is very strict. It's very detailed. But among them, it's very detailed. I'm also careful about the code quality. Even if I think I wrote a good code, I get such comments. As an engineer, I have a wide view. I have a lot of opportunities to learn new things and new concepts. Especially in our team, we have a team that adopts Elixir. We have a team that commits to Elixir and has a member who knows Elixir. I think it's a great opportunity to learn minor languages. I think it's a great point to work with such an excellent member. The third point is that you can commit to the list and business of an engineer. In a sense, an engineer is not just writing code. For example, you can say that you should improve the idea of the function. Or you can say that you should quit the company you want to work with. I think that's beyond the level of other companies' engineers. The list has developed with such opinions as a team. In a sense, it's not just an engineer. You can make a business with us. I think it's a great opportunity to work with a big IT company that only does coding. If you're tired of it, I think it's fun. I think so too. I think these three points are really good.
Ryohei Watanabe: Thank you very much.
Ryohei Watanabe: The last question is, if you were to start a company, do you have any advice?
Masashi Yokota: I think there are two perspectives. I think it depends on whether you want to be a CEO or a technical person. In terms of CEO, you can't make a business by yourself. So you have to find a good member. I think that's one of the points. Another important thing is that you have to invite people from different cultures. So you have to talk to each other with an open mind from a early stage. I think it's most important to know how each other thinks. I think there are a lot of fights at first. In that sense, it's not good to think that you were thinking about something big. So I think it's important to fight a lot in the beginning and understand each other. As a CEO, you have to have a broad mind. I think it's a very good CEO because you can do that. Next question is for people who want to be the first engineer. In this pattern, I don't have a business idea, but I want to challenge if there is something interesting. This is my opinion. I have a position talk, but I think a niche business is good. Even if you talk to your parents or friends, no one will understand. But I think it's the best business. This reason came from my experience. When I was in college, I was making a language AI. At that time, the boom was AlphaGo. It was a language AI that beat the top pro-agents. I couldn't believe it at that time. My project was to make a Japanese-made AlphaGo and make a language AI that is stronger than AlphaGo. It was really reckless to compete with Google with all the skills, engineers, and money. I was able to beat the top pro-agents, but I couldn't get the quality to challenge AlphaGo. At that time, I thought I shouldn't fight with the giant tech. I think it's easy to understand that they will find it. I thought I should stop if they were interested in joining. I think it's a good business. I think it's a good business. I think Google will find it if it's something that you think of. Google will never know our business.
Ryohei Watanabe: I think it's a great answer. Do you have any advice for people who want to apply for RESTAR?
Masashi Yokota: I talked a little bit about it earlier. As a team, we always want professional members. In a sense, I have to be able to appeal to people who are strong in this field. I think it's important to understand basic knowledge such as OOP and FP. I don't think we ask difficult questions. But I think it's difficult to answer. I think it's important to catch up with the technical concept you know.
Ryohei Watanabe: I think it's important to catch up with the technical concept you know. I think it's a good answer. Thank you very much.