Faculty of Science and Engineering, Doshisha University

#NAOqi #ComputerAssistedLanguageLearning

I developed a system that allows learners to practice English expressions by interacting with two robots. I conducted an experiment targeting native Japanese-speaking students, manipulating the questions and model answers, and analyzed the learners' audio responses and eye movements. I created 30 scenarios, each about 10 minutes long, to realize efficient learning. Received the Student Encouragement Award at the IPSJ (Information Processing Society of Japan) National Convention.

Wellbeing Research Center, Doshisha University

#DataScience #BehavioralChange

Participated in an industry-academia collaboration project aimed at improving productivity in telework environments during the COVID-19 pandemic. I was responsible for everything from survey analysis to the development of an app that promotes behavioral change, addressing new challenges brought about by lifestyle shifts.

Graduate School of Business, Doshisha University

#GlobalBusinessManagement #Coaching

While studying business administration and organizational management at the graduate school as a research student since my undergraduate years, I conducted research on organizational reform within the Self-Defense Forces.
Additionally, aiming to promote mutual learning through academic studies, I established a community for practicing coaching and engaged in activities with its members. Under the guidance of experts, we gathered over 180 members and ran activities for two years. (Doshisha Coaching Lab)
As part of our activities, we conducted a career survey among 3rd-year undergraduate students. Finding that over 70% of the students could not clearly envision their career plans gave me a sense of crisis. Wanting to create an opportunity not only for students but also for working professionals to gain new perspectives, I planned an event inviting corporate executives to engage in dialogue with students.

Global Leader Development Program

#ResourceManagement #PublicPolicy

Focusing on the increase in power demand accompanying the expansion of AI utilization, I researched environmental and energy policies. I visited energy facilities and government agencies in Europe to investigate the initiatives and challenges in each region. I also proposed my own plan for achieving sustainability by 2050.

Graduate School of Engineering, The University of Tokyo

#AI #LearningTracking

Using a tablet learning tool specifically designed to develop basic academic skills, I developed a machine learning model to improve learning efficiency and predict the forgetting curve of elementary and junior high school students. The model highly accurately predicts the timing when students are about to forget kanji, English vocabulary, or calculations, and assigns homework just before they forget. This achieved an optimal review schedule tailored to the unique characteristics of each student.

Volunteer Activities

I believe I have accumulated experiences worth passing on to the next generation.
I look forward to an expansive and bright future with all of you.

Activities at YMCA

#CommunityContribution #InternationalExchange

I belonged to the YMCA, an international youth organization, for over 10 years starting from age 13. I put particular effort into community exchange, participating in various events and programs. During junior high and high school, I contributed to community development through neighborhood cleanups and fundraising activities in Kyoto City. After entering university, as a program leader at ski camps, I provided ski instruction and managed recreation for students from elementary to high school. I tried to foster a positive attitude in children by promptly finding and praising even small accomplishments. I believe that an environment where we can acknowledge and uplift each other nurtured the current version of me—someone who can think independently and take proactive actions. I firmly believe that the inclusion of diverse elements such as nationality, race, gender, and age brings diverse perspectives, which ultimately leads to unexpected discoveries and collaborations.

---

Establishing a Medical Supplies Support Organization

#ReliefSupplies

I established a volunteer organization to match medical institutions and welfare facilities with people wishing to donate, thereby supporting their contributions. During the state of emergency in March 2020, hospitals in Kyoto City were consuming as many as 300 disposable masks a day, and support could not keep up with the demand. Along with 56 classmates, I visited various facilities in the city to collect raw information from the frontline and strengthened our outreach through Facebook and Teams. Furthermore, we published tutorials on how to make handmade masks and disposable gowns, and collaborated with city council members to arrange the rental of sewing machines from civic halls. As a result, we acted as proxies to collect contributions from about 200 individuals who had felt helpless not knowing how to assist, successfully delivering supplies to 7 organizations within a short period of 3 months.

Healthcare Industry Initiatives

2023
Realization of Rentable yet Non-Transferable Personal NFTs
: Building a Transparent Data Management System for Medical Data Utilization

Background

In the contemporary medical field, data utilization has yet to advance sufficiently. I focused on converting data managed by medical institutions into "data that individuals can recognize and utilize as valuable." However, this concept required the construction of a mechanism that ensures both security and transparency. There had been no successful implementation of such a concept, and many technical and operational challenges remained unresolved.

Challenge

While individuals providing health data wish to limit the elements and duration of the data they share, data users such as insurance companies seek a system that guarantees data accuracy while enabling its use for business purposes.

Action

Individual development:
1. Integrated the technologies of "rentable NFTs" and "non-transferable NFTs," which had previously been independent, to develop NFTs that respond to diverse needs.
2. Built a VC issuance system using "Blockcerts" based on the Verifiable Credential Data Model specification.
3. Designed and developed a system for transferring tokens from users to data holders through health data rental.

2024
Predicting Clinical Trial Success of Immune Checkpoint Inhibitors
: AI Utilization for Efficient and Effective Drug Discovery

Background

Drug development takes more than 10 years and costs hundreds of billions to trillions of yen, with the success probability declining from 1 in 15,000 ten years ago to 1 in 30,000 today. Under these circumstances, optimizing and accelerating clinical trials and resource allocation has been a long-standing challenge for the industry. In particular, the success rate for transitioning from Phase II to Phase III is only 24.6%, leaving significant room for improvement.

Challenge

Worked on predicting the probability of success in Phase III clinical trials for immune checkpoint inhibitors. Applied the concept of "acquiring representations from observational information" from world models, and visualized and verified improvements in accuracy based on patient blood data and vital data.

Action

Team development.
1. Practice of the process of gradually transforming the abstract concept of the immune environment into specific content.
2. Implementation of AI that directly addresses social issues through medical projects expected to undergo major changes in the next 10 years.

2025
Emotion Estimation × Educational Support AI "EMOTIP"
Exhibition at Expo 2025 Osaka, Kansai
: Improving Care Quality and Reducing Turnover Rates at Care Facilities Using AI Video Analysis

Background

The care industry faces a chronic labor shortage. A survey of more than 500 care workers revealed that "no future prospects" at 15.4% and "current salary" at 14.9% were among the top reasons for leaving within three years, confirming that anxiety about the future and working conditions are the main factors driving turnover.
Furthermore, the difficulty in communicating with residents who have difficulty expressing emotions, such as those with dementia, was identified as a major source of stress for both residents and caregivers.

Challenge

There is a need for a system that correctly understands the emotions of care recipients and provides appropriate support, thereby reducing stress for residents and the burden on caregivers. Rather than traditional observation and record-based assessments, it was necessary to visualize care quality and support staff in self-evaluating their care skills, reducing the burden of staff education and lowering turnover rates.

Action

Exhibited at the 2025 Japan World Exposition (Expo 2025 Osaka, Kansai) "Experience of Future Living" (July 8–14), promoting a project to share everyone's vision for the future with the world through the "Tree of Life Shining Future."

1. Developed a system that uses small far-infrared cameras and a proprietary large language model (LLM) to analyze video with high accuracy, analyzing the emotional states and behavioral patterns of care recipients.
2. Integrated a feedback function that quantifies and visualizes the quality of caregiver responses based on AI analysis results, as well as a reassurance monitoring function that shares daily living conditions and "how much enjoyment residents are experiencing" with family members.