About

Summary

I work as a Unity engineer and director in mobile game development, and as an ML engineer for speech synthesis (TTS) building AI characters and voice dialogue services. I publish Japanese TTS and Unity libraries as open source, and release models and datasets on Hugging Face.

Experience (by role)

• Unity client engineer, mobile games — PC and smartphone clients for IP titles, core libraries, and engineering and team building for a new business in its launch phase
• Unity engineer, xR — VR/MR apps for B2B and B2B2C, and metaverse projects
• Contract work — voice dialogue services, AI characters (TTS), datasets and training for speech and singing voice synthesis, and Unity development of 3D spaces for an education service

Skills

• Speech AI — TTS pre-training and fine-tuning, Japanese G2P and pitch accent, adapting streaming ASR, turn detection and spoken dialogue models to Japanese, and on-device inference with ONNX and similar runtimes
• Unity / C# — client development, library design, CI (GameCI, GitHub Actions)
• AI characters — characters that combine speech synthesis, LLMs and voice input

Activities

• Organizer of “Generative AI Nandemo Exhibition”, an in-person exhibition for creators who make things with generative AI (planning and operations; held since 2024 through vol.6, with over 5,000 visitors in total)
• Running the MidraLab developer community
• Indie game development