Interview with AI Engineers

What kind of work are you both currently involved in?

It sounds so difficult.
What specific tasks are involved in improving the accuracy of skeleton detection through additional training?

Basically, we create Annotation data focusing on areas where recognition is weak, and train the model using it. Additionally, we conduct data processing as needed and proceed with the training.

When I was working as a part-time employee, I also did annotation data creation! It’s quite detailed work, isn’t it?

There are times when it’s unclear whether the accuracy will improve as expected, so it requires not only knowledge but also a sense to efficiently train the data.

How about Mr. Aoyama? What kind of work are you involved in?

I’m involved in creating data for machine learning, selecting algorithms, evaluating analysis results, and developing tools to make machine learning more user-friendly.

This also seems difficult!
What kind of tools are there to make machine learning easier to handle?

For example, there are viewers (software for viewing the contents of files) that visually display Training data in an easy-to-understand way. We also develop applications that utilize data obtained from machine learning for various purposes.

What were the reasons behind the interest in Artificial Intelligence (AI)?

I was interested in AI because I studied it at university. During my university days, I focused on the idea of AI complementing human capabilities by imagining what lies beyond human hands, which has proven incredibly useful in my current job at the company.

My last job was too far from AI-related work, haha.

After graduating from a shoe design school, I worked for a company involved in planning and manufacturing shoes. Then, due to transfer, I came to Fukuoka.

Shoes!?

How did you become interested in AI from there?

The trigger was when I saw a TV program about AI one day, and I thought, ‘Wow, AI seems really interesting.’ From that moment, I decided to quit my job and study AI. Since I’ve always liked mathematics, learning wasn’t too difficult for me. After a few months of self-study, I became proficient in coding with Python.

Why did you think of joining NEXT-SYSTEM?

I’m originally from Fukuoka, but I went to university in the Kansai region. However, due to various reasons, I ended up returning to Fukuoka. When I looked into AI companies in Fukuoka, I came across Next System. The deciding factor was, of course, my interest in AI. Also, the fact that the CEO has an engineering background was significant.

In my case, while researching various AI companies, I found that they specialized in image recognition, which is what I wanted to pursue. Also I was particularly attracted to the fact they were dealing with skeleton detection since I love sports.

What do you think after joining this company?

Working for a company that utilizes cutting-edge technology allows me to constantly engage with new information and techniques, which is an experience that’s hard to come by through self-study alone. Additionally, I struggled a bit when I first joined the company because the level of other engineers were so high. I spent my time outside of work studying diligently every day.

Truly, being surrounded by such high-level peers has allowed me to grow technically.

Also, the field of behavior analysis that I’m currently working on is incredibly fascinating, and I feel fulfilled everyday. Especially, the recognition prediction in VisionPose is connected to the research I did during my university days. So I feel like there might not be a company in Japan that fits me better than this one!😄

That’s right, it’s because of the environment that allows for such growth and potential for advancement that we’re where we are today.

Do you think that the strength of NEXT-SYSTEM lies in its engineers? What do you think are NEXT-SYSTEM’s strengths?

I think the level of engineers is also a strength. That’s what led to the development of products like VisionPose.

VisionPose is truly an outstanding product. It offers high flexibility as it allows for the addition of key points and reinforcement of weak postures depending on the consultation.

Exactly, having many experienced engineers significantly accelerates our learning speed as young employees.

In addition, the fact that our CEO is an engineer is also a major strength. It’s extremely helpful to have an environment that is suitable for engineers. Furthermore, the CEO’s ability to provide technical explanations to clients is another strength of this company.

I believe that both of you will continue to be key person supporting NEXT-SYSTEM. Do you have any goals or aspirations for the future?

My goal is to create and implement open-source projects. As an engineer, there’s nothing more fulfilling than having something I created used by as many people as possible. I hope to able to provide technology that can contribute to society, even if it’s not just open source.

I want to be more proactive in learning AI and applying it to new things. First, by gaining solid technical skills and knowledge, I aim to be able to offer insights to other engineers around me. I believe this will contribute to the company as well.

Lastly, any suggestions you want to give who is aiming to become an AI engineer or considering joining an AI company?

AI has endless possibilities depending on the ideas. It’s truly fascinating, and I want more people to know about this infinite potential.

Every day, as I research new AI fields, I sometimes wonder, Is this really surpassing human capabilities? haha

It’s recommended for those who have a desire to take on challenges or pursue their passions, particularly in utilizing cutting-edge technology. I personally had a strong interest in AI when I first joined the company, and since then, I have been fortunate to work on various AI-related projects.