About

Mao Saeki / 佐伯真於

I am a dialog system researcher at Equmenopolis Inc., where I lead the development of the InteLLA virtual agent. I am also pursuing a Ph.D. in Computer Science at Waseda University under the supervision of Prof. Tetsunori Kobayashi. My interest lies in multimodal conversational AI, especially the understanding and generation of non-verbal cues. I have built AI agents from the ground up and scaled them to serve thousands of users, while conducting cutting-edge research on turn-taking and gesture generation, integrating them into real-world applications.

Projects

InteteLLA - Intelligent Language Learning Assistant

InteLLA automatically measures the conversational profiency of English learners through naturalistic conversation, enabled by gesture generation, turn-taking and proficiency assessment models. By adapting the conversational difficulty to each user though real time assessment, it is able to extract user's full potential, and give accurate assessment. InteLLA has previosly won many research awards.

We are actively looking for collaboraters! Contact me if you are interested.

Experiences

  • Equmenopolis Inc., Research Scientist, 2022 - Present
  • Waseda University, Research Associate, Perceptual Computing Group, 2020 - 2023
    Developed conversational AI agent for language learning assistance
  • National Institute of Advanced Industrial Science and Technology, Research Assistant, 2018 - 2019
    Worked on anomaly detection for industrial machinery
  • LPixel Inc., Development Intern, 2018 - 2019
    Worked on lung cancer detection from CT scan images

Education

  • Ph.D. in Computer Science and Engineering, Waseda University, 2020 - Present (Advisor: Tetsunori Kobayashi)
  • M.E. in Computer Science and Engineering, Waseda University, 2020 (Advisor: Tetsunori Kobayashi)
  • B.E. in Mechanical Engineering, Waseda University, 2018

Awards

  • 2025 Telecom Interdisciplinary Research Award, Telecommunications Advancement Foundation, March 2026
  • Best Paper, SIGdial, September 2024
  • Best Student Paper, Interspeech, August 2023
  • Best presentation, The Japanese Society for Artificial Intelligence, June 2022
  • Reimagine Education Award, Learning Assessment Category Bronze, Quarrelli Simmons (QS) and The Wharton School of the University of Pennsylvania (MBA), December 2021
  • Young Researcher Award for Excellent Research, The Japanese Society for Artificial Intelligence SIG-SLUD (Special Interest Group of Speech, Language Understanding and Discourse Processing), October 2021

Publications

  • Ryuichi Sumida, Mao Saeki, Masaki Eguchi, Sadahiro Yoshikawa, Koji Inoue, Tatsuya Kawahara, Yoichi Matsuyama, "Memory-Driven Self-Disclosure and Relational Turning Points: A Longitudinal Multimodal Study of Human-AI Interaction", Proc. ACM International Conference on Multimodal Interaction (ICMI), 2026.
  • Mao Saeki, Masaki Eguchi, Yoichi Matsuyama, and Tetsunori Kobayashi, "Eliciting User Interaction through the Incremental Confirmation of User Understanding Using Multimodal Cues", Dialogue & Discourse, vol. 17, no. 2, pp. 53--93, 2026.
  • Yuto Abe, Mao Saeki, Atsumoto Ohashi, Shinnosuke Takamichi, Shinya Fujie, Tetsunori Kobayashi, Tetsuji Ogawa, Ryuichiro Higashinaka, "Effects of Dialogue Corpora Properties on Fine-Tuning a Moshi-Based Spoken Dialogue Model", Proceedings of the 16th International Workshop on Spoken Dialogue System Technology (IWSDS), pp. 104-108, 2026.
  • Fuma Kurata, Mao Saeki, Masaki Eguchi, Shungo Suzuki, Hiroaki Takatsu, Yoichi Matsuyama, "Development and validation of engagement and rapport scales for evaluating user experience in multimodal dialogue systems", arXiv preprint arXiv:2505.17075, 2025.
  • Mao Saeki, Hiroaki Takatsu, Fuma Kurata, Shungo Suzuki, Masaki Eguchi, Ryuki Matsuura, Kotaro Takizawa, Sadahiro Yoshikawa, Yoichi Matsuyama, "InteLLA: Intelligent Language Learning Assistant for Assessing Language Proficiency through Interviews and Roleplays", Proc. The 25th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGdial), September 2024.
  • Fuma Kurata, Mao Saeki, Shinya Fujie, and Yoichi Matsuyama, "Multimodal Turn-Taking Model Using Visual Cues for End-of-Utterance Prediction in Spoken Dialogue Systems", Proc. The 24th Annual Conference of the International Speech Communication Association (INTERSPEECH), August 2023.
  • Mao Saeki, Kotoka Miyagi, Shinya Fujie, Shungo Suzuki, Tetsuji Ogawa, Tetsunori Kobayashi, and Yoichi Matsuyama, "Confusion detection for adaptive conversational strategies of an oral proficiency assessment interview agent", Proc. The 23rd Annual Conference of the International Speech Communication Association (INTERSPEECH), September 2022.
  • Mao Saeki, Ryuki Matsuura, Shungo Suzuki, Kotoka Miyagi, Tetsunori Kobayashi, and Yoichi Matsuyama, "InteLLA: A Speaking Proficiency Assessment Conversational Agent with Adaptive Interview Strategy", The Japanese Society for Artificial Intelligence (JSAI), SIG-SLUD, October 2021.
  • Mao Saeki, Weronika Demkow, Tetsunori Kobayashi, and Yoichi Matsuyama, “A WoZ Study for an Incremental Proficiency Scoring Interview Agent Eliciting Ratable Samples“, Proc. The 12th International Workshop on Spoken Dialog System Technology (IWSDS 2021), November 2021
  • Mao Saeki, Yoichi Matsuyama, Satoshi Kobashikawa, Tetsuji Ogawa, and Tetsunori Kobayashi, “Analysis of Multimodal Features for Speaking Proficiency Scoring in An Interview Dialogue“, Proc. The 8th IEEE Spoken Language Technology Workshop (SLT 2021), pp.629-635, January 2021.

Contact

email: saeki[at]equ.ai