Gugan Kailasam, serving as an Assistant Manager and AI Technical Architect at Hitachi Solutions, is on a mission to deliver superior AI solutions to the world. Following his relocation from India to Japan in the summer of 2024, Gugan is leveraging the collaborative strengths of both nations to promote AI technology and its applications. 

Bridge-building between India and Japan

── You’ve been involved in AI projects  in Hitachi Solutions India since 2020. Could you describe your specific role?

As an Assistant Manager and AI Technical Architect at Hitachi Solutions India, I support the AI and R&D projects commissioned by our Japanese headquarters, Hitachi Solutions. My role covers three main areas. First, I facilitate project execution by serving as a direct channel of communication between Hitachi Solutions and Hitachi Solutions India, translating concepts and sharing technical insights. Second, I serve as the primary liaison between our various business units, addressing project-related queries and disseminating information about the accomplishments of Hitachi Solutions India’s R&D center to Hitachi Solutions. Third, I provide specific support for launching new AI-based businesses and developing AI solutions for customers, including, for example, proposal creation.

── You recently relocated to Japan. What motivated this move?

The primary objective of my relocation to Japan is to enhance the visibility of Hitachi Solutions India’s AI technology and team capabilities. Now that I am on-site, I can effectively communicate with various stakeholders, showcasing the expertise of our Indian team members and demonstrating how we can support Hitachi Solutions in executing customer projects. 

Given the ongoing rapid advancements in generative AI, it is challenging for Hitachi Solutions members to stay abreast of all the pertinent developments by themselves and provide optimal solutions to customers. Our team aims to alleviate the load by offering consultation and technical support on cutting-edge technologies. Sharing this knowledge between Hitachi Solutions’s business units is a key aspect of my mission.

Additionally, I engage in regular discussions with Hitachi Solutions members in order to understand their requirements and advise them on project feasibility. I leverage our past experiences and expertise to provide insights on potential approaches, facilitating smooth collaboration between Hitachi Solutions and our offshore delivery team.

── So, you are a bridge between the Indian team and Japan?

Yes. As a team leader within Hitachi Solutions India, I communicate the Japanese team’s expectations to our Indian colleagues. I help them to navigate cultural nuances, in particular, for example, when our Japanese counterparts might not express all their thoughts directly, and make our team mindful of any implicit expectations.

Furthermore, given the rapidly evolving AI landscape, I ensure that our team stays engaged in continuous learning. I identify crucial topics for skill development and organize knowledge-sharing sessions to keep the team updated on the latest technologies. To keep myself informed, I participate in online courses and review the most recent research papers on platforms like X and LinkedIn.

Showcasing AI’s Potential

── What kind of work did you originally do?

In my career to date, I’ve been involved in over 20 major projects, primarily focusing on AI and various product research and development initiatives. Notable projects include power consumption and people inflow prediction for building management. We have also conducted research on identifying minute objects in video or images, which has potential applications in construction and other industries where cameras are positioned at a distance from small objects.

── Tell us about some of your most impressive projects.

The first project I worked on was in 2020, which pertained to a system called employee churn prediction. We built a proof of concept (POC) for a system where AI predicts whether or not an employee will quit an organization based on certain information about the employee.

For example, if you were to input that I have worked with Hitachi for a certain number of years, along with various other information about myself, the system would predict whether or not I will quit the organization. This kind of information helps to identify the factors that influence an employee’s decision to leave an organization.

During our research, we found that factors such as extended working hours and irregular schedules contributed to employee turnover. Based on these findings, we developed a system to help the HR team identify and address issues that could lead to employees leaving. It’s important to note that this project was conducted before the advent of generative AI.


── Are the insights you gained then still useful today, in the era of generative AI?

Although the technology at that time was clearly different, as it was not generative AI, there are similarities in terms of the insights we gleaned and the manner in which we assessed the data. Data provides some kind of information that a model understands and then does something in response to. 

In the case of employee churn prediction, a model understands the employee information and then predicts whether or not the employee will quit. Likewise, in the case of text generation, the model understands a question from a user and generates answers. The input to these kinds of AI models is data. Understanding the data, how we analyze it, and determining what data will be useful for a particular scenario – these aspects of data exploration and evaluation are common to all projects.

Knowledge acquired from training an AI model can also be applied to various different projects. So, what we learn in one project can naturally be used in other projects, too. There is also the practice of fine-tuning: we build a model, gather the data, train the model, and then we fine-tune it. There are always various parameters at play. It’s rather like all the knobs you see in an airplane cockpit. There will be a certain number of knobs in any model, and we practice a kind of tuning: if you adjust this “knob”, the model will behave in this way. As AI data scientists, we tune these knobs to improve the AI model’s performance. This tuning process can most certainly be replicated or streamlined and applied to other projects.

── What do you enjoy about your current job?

My current role will broaden my perspective on AI. Previously, I viewed AI from a more technical perspective. Now, I consider its practical applications in solving operational challenges for our customers, its potential to create new value, and its societal impact, and I’m developing the skills to build bridges between AI’s vast untapped potential and real-world business applications, which is a challenge that I find particularly engaging.

── The emergence of generative AI, especially interactive systems like ChatGPT, has significantly altered people’s understanding of AI. What are your thoughts on this?

You’re right. The advent of ChatGPT and similar technologies has had an enormous impact on public understanding of generative AI. People can now more easily grasp its capabilities and potential applications. That said, there remains a significant lack of know-how in terms of fully comprehending how to leverage this technology to address specific problems. This is where our consultative role is indispenable. When clients present business challenges, we can advise them on potential generative AI solutions.

Conversely, we sometimes encounter situations where clients have unrealistic expectations of AI. It’s vital to provide a credible assessment of what’s achievable. We often come across knowledge gaps of this kind, which I then help to address through direct discussions with customers.

Our role is not simply to showcase AI’s possibilities, but also includes setting realistic expectations and identifying practical, implementable solutions tailored to each client’s unique needs.

Spreading AI Awareness

── What are some new challenges you would like to work on?

I’d like to interact more with customers in Japan. This is something that I didn’t have much experience of while working in India, as I was primarily engaged with the business team there. I haven’t had the chance to directly interact face-to-face with customers and build relationships with them, and I believe that building strong relationships with customers is crucial for any business.

As an Indian who knows Japanese, I’m curious to see how customers here will react to me and how they will perceive me. I want to try to support them and solve their problems directly, and to see if I can actually build good relationships with them. This is something that I look forward to discovering first-hand: how I’ll be received as an Indian professional in the Japanese business context.

── What synergies can Hitachi Solutions expect from working with the Indian team?

India’s offshore business model exposes us to a global clientele, including customers from Japan, America, Europe, and various other regions. The flexibility and diversity of approach that this requires certainly keeps us on our toes. In particular, the American and other international markets often actively seek out cutting-edge technologies, which pushes us to continuously evolve and stay competitive on a global scale, not just within India.

While Japan isn’t necessarily lagging behind in this regard, there may be something of a disparity in terms of the level of familiarity with AI among business professionals in general. There are so many different industries in which AI can play a role, and while generative AI is currently in the spotlight, it’s crucial to recognize AI’s broader potential. A deeper understanding of AI’s capabilities across different domains would greatly benefit businesses in navigating this technological landscape more effectively.

── You’re working to expand AI’s reach throughout society via Japanese-Indian collaboration. What broader social impact do you envision?

In the business world, the arrival of generative AI has been the catalyst for significant changes. While AI won’t entirely replace jobs, it will certainly result in the automation of certain activities, particularly routine, less knowledge-intensive tasks. This shift will allow humans to focus more on value-added, knowledge-intensive, and creative work. I see this as the emerging trend.

What is more, AI is also increasingly empowering the general public by making truly enormous troves of online information more accessible. The real societal change will come from how people leverage such AI-powered knowledge in their daily lives and work.

AI is democratizing access to advanced knowledge. This accessibility enables the general public to learn more, do more, and in turn, contribute more to society. My hope is that we can bring deeper awareness of AI to a broader audience, fostering a more informed and capable society.

Personal Perspectives

When I’m not working, I love staying active. Basketball is my go-to sport, but I also enjoy playing cricket and going for runs. The other day, I climbed Mt. Takao. The weather was nice, and the view from the top was wonderful. On the mountain path, I enjoyed a pleasant hike surrounded by nature and lush greenery. I also savored some delicious dango (rice dumplings) on the way up, which added to the fun. Climbing the mountain was refreshing. The view of the Tokyo cityscape and the distant mountains from the summit was spectacular. It was a truly valuable experience to spend time in a place surrounded by nature, which is also beneficial for one’s health. After we descended, we had delicious soba noodles, making it a very satisfying day.