Ph.D., NAIST Best Student Award
2017 - 2019
Information Science
M.Eng.
2015 - 2017
Information Science
GPA: 2.96/3.00
B.Sc., cum laude
2010 - 2014
Informatics
GPA: 3.50/4.00
Dialog Systems and Machine Learning, Heinrich Heine University, Germany|May 2019 - present
Unsupervised Dialogue Policy Optimization: Developed LAVA (Latent Action Spaces via Variational Auto-Encoding), optimization frameworks, and reinforcement learning strategies to train dialogue agents without explicitly predefined action spaces.
Large Language Models, Calibration, and Hallucination: Investigated learning and generation behaviors in LLMs for task-oriented dialogue, directly investigating how models manage uncertainty, out-of-distribution data, and systemic errors across the training lifecycle.
Academic Teaching and Mentorship: Jointly designed and delivered graduate-level curricula on Reinforcement Learning and Spoken Dialogue Systems; mentored and supervised dozens of Bachelor's and Master's thesis projects in NLP and ML.
SCOPE Project, Nara Institute of Science and Technology, Japan | October 2016 - 2018
Affect-Sensitive Dialogue Generation: Designed and evaluated end-to-end neural network architectures capable of generating emotionally aware chat responses, focusing on positive emotion elicitation in conversational agents.
Statistical Modeling & Multi-Modal Frameworks: Built statistical frameworks to track, understand, and predict emotional trajectories and conversation dynamics in complex human-to-computer interactions.
Honda Research Institute Japan, Japan | July - September 2015
Multi-Modal Emotion Recognition: Engineered machine learning models for joint audio-visual emotion recognition utilizing spontaneous, natural conversational data in Japanese.
Pipeline Development: Designed specialized data-collection pipelines and feature-extraction protocols optimized for multi-modal speech and visual affect analysis.
Augmented Human Communication Laboratory, Nara Institute Of Science and Technology, Japan | June - August 2013
Social-Affective Computing: Conducted deep analytical studies on the automatic prediction of emotional triggers and behavioral responses within spontaneous conversational datasets.
Heinrich Heine University Düsseldorf
December 2022
21st Annual Meeting of the Special Interest Group on Discourse and Dialogue
July 2020
Nara Institute of Science and Technology
March 2019
The Japanese Ministry of Education, Culture, Sports, Science and Technology
October 2014 - March 2019
The Technological Advancement Foundation
April 2017
IEEE Spoken Language Technology Committee
December 2014, South Lake Tahoe, United States of America
International Committee for the Co-ordination and Standardization of Speech Databases and Assessment Techniques
October 2014, Phuket, Thailand
Japan Student Services Organization
June - August 2013
Bank Indonesia
July 2007 - June 2010
Python, C++
Deep Learning on PyTorch and Tensorflow
Python toolkits and libraries, e.g. numpy, scipy, nltk, etc.
Proficient in working on large clusters and Google Cloud Platform
Native
Proficient
TOEIC: 985/990
TOEFL:113/120
Intermediate proficiency
B2
Elementary
JLPT: N5 157/180
I've been practicing fiber arts almost daily since 2017, mostly knitting and spinning. I have a huge appreciation for natural fibers, especially wool.
I play the ukulele from time to time.
I love traveling and photography.