Message for students 2020: Keiichi Kisamori
February 7, 2020
Promoting AI and simulation in collaboration with AIST
Assistant Manager at Data Science Research Laboratories
Kisamori joined NEC after a one-year post-doc experience. He is involved in promoting a combination of AI and simulation research in a joint laboratory with the National Institute of Advanced Industrial Science and Technology (AIST). His hobby is watching soccer matches, and he is active in cheering on J League teams.
From physics research in academia to machine learning research at a company
I originally studied nuclear physics at university. I worked as a post-doc for one year at an R&D group after obtaining my doctorate, however, I began to realize that I wanted to experience the world outside academia. Although I enjoyed physics research and experimentation and there was talk of a good post, I had a strong desire to go out into the world instead of just remaining in academia. Just around that time I began to think that machine learning was interesting, and I wanted to do corporate research in a different field.
I discovered NEC when I decided to go out on a whim and attend a joint corporate briefing event. Although I met a number of people from various companies, I felt an overwhelming sense of something in common with NEC personnel. Perhaps it speaks to my qualities as a researcher. I was very drawn to something about NEC and I decided to apply. Later, Fujimaki-san, who would turn out to be my boss, was my interviewer. Although he showed up in a T-shirt and bright green hoodie, he began firing apt, pointed questions at me one after the other. The conversation took a turn into a world view that I had never though of in academia, and because it seemed like such a fun place to work, I decided to join NEC. So for me, the final deciding factor in joining was probably that I was drawn to the people who worked there.
In my team there are people in mathematics, those who involved in experimental physics, and those engaged in bio-related research, forming a solid, wide-ranging backbone. In machine learning research, I don't believe there is any need to limit one's possibilities in joining a company just because your field of specialty is different.
Greedily enjoying the good from both corporate research and academia
I am currently engaged in research at the academia-industry laboratory launched through the collaborative efforts of NEC and AIST (National Institute of Advanced Industrial Science and Technology). We are building a digital twin hinging on technology that incorporates simulation technology with in AI in order to search for all possibilities and patterns, and a major research theme is aiming to optimize the business developed by companies that introduce this technology.
Corporate researchers must show their own technological value, so dialog with the customer is very important. You must listen to fundamental requests for commercialization, then ask how much your research and technology can be used to solve such issues. Also how much money can be made in such efforts? The time spent discussing and thinking over these elements with the customer is very refreshing and fun. In the end, I believe I want to do something that makes others happy. It really is great to hear a customer say "this technology is amazing!" Although one of the final goals of the corporate researcher is to make money, I think that is something that comes afterward. The most important thing is providing technology that helps customers out of a bind or makes them happy. This is the kind of joy that only a corporate researcher can experience.
That being said, the opportunity to be active in academic conferences is also part of NEC's appeal. NEC is one of the few major Japanese corporations whose papers are accepted at global top level conferences. NEC also encourages us to participate in conferences. One of the holdovers in our team from when Fujimaki-san was the leader is the significant corporate contribution that raising NEC's presence brings when producing outstanding achievements at conferences. Actually just the other day my research paper was accepted at AISTATS*, a world-class AI-related academic conference. I believe that one of the great things about NEC is the chance to greedily enjoy both the best of corporate research and academia.
- *International Conference on Artificial Intelligence and Statistics
A culture of challenge and freedom
For now, my personal goal is to complete a technology that I started from scratch, which finally looks like it will be commercialized. However, I am greedy, so I don't want it to end here (laughs). Actually I'm currently thinking bigger, and discussing plans with the general managers. I am utilizing the framework in place through collaboration with AIST to conceive of a plan to provide an attractive place for researchers.
This concept is part of it, but the challenge makes work more interesting. I am a risk taker, so I look forward to taking on many challenges that interest me, regardless of the danger. Perhaps it is this part of my personality that makes this job seem more like a hobby, allowing me to truly enjoy myself whether I am at work or not. NEC rises to such challenges, respecting each individual employee and providing employees with the freedom and discretion to make choices.
There is also great freedom in how we're allowed to work. The super flex system eliminates any set times for coming and going and concept of core time, and employees are allowed to select their preferred work location and even telework. I love the shared office that NEC rents, and spend several days a week there. I also have a desk at AIST, so I am only at my NEC desk probably once a week (laughs). But with all that freedom comes responsibility. You must pursue solid results with all sincerity. But even so, there is no reason to fear. I would love to work with people who like to take on big challenges. The more people that come work here, the more fun our research will be.
A day at work
Message to my past self in my school days