Justin Turkat, Founder and Managing Partner of SIP Capital, said about the company that it is “trailblazing the future of voice AI, right here in Japan. We’ve backed Nori, Jungo and the team since 2023 and are excited to continue the journey.”
Fugaku, GENIAC, and DOTSU
The company’s origins trace back to a project focused on developing a Japanese-centric large language model (LLM) using the supercomputer Fugaku. Building on this foundation, Kojima and Kasai assembled a team of researchers and engineers to enhance generative AI’s speech capabilities.
In 2024, the company released the beta version of its “Kotoba SpeechGen,” which uses preset voices and voice cloning to produce fluent and natural voice generation. DOTSU iOS app followed in March 2025 with speech-to-speech translation along with transcription. The app has quickly gained traction, logging over 500,000 sessions in three months.
The company started rolling out APIs for automatic speech recognition (ASR) and other voice agent components to enterprises and developers in late July 2025. Initial use cases are remote business meetings, soon to expand to entertainment and events, a company representative shared with Slator.
Kotoba Technologies was also selected twice for Japan government’s “GENIAC” generative AI project, which leverages large-scale computational resources to advance speech AI infrastructure.
Looking ahead, the startup plans to expand beyond the US and Japan, along the way increasing language availability to Chinese, Korean, and Spanish. While the company’s primary focus is on the B2B market, its DOTSU iOS app is also suitable for consumers. A desktop app for speech-to-speech translation is also expected soon.
The newly secured capital will be used to accelerate the transition from R&D to commercialization. The company’s immediate plan is to invest in hiring machine-learning and application engineers to advance model development and productization. This move is intended to quickly bring its speech-AI technology to a wider market.
