Anime, manga, and games, with a take · A Yukimedia publication

Kotoba to Supply Voice AI Models to Cresta's Contact Center Platform

Kotoba Technologies, the San Francisco company with an R&D base in Chiyoda, Tokyo, plans to partner with California-based Cresta and supply its speech recognition and speech synthesis models to Cresta's Voice AI agents for contact centers.

Reporting from 1 source: ASCII.jp.

Why it matters

Kotoba is selling its Japanese-language voice models through a CX platform that already serves major companies, which puts the hard part of the problem, honorifics, homophones, phone-line noise, and turn-taking, in front of enterprise callers rather than a demo audience.

Kotoba to Supply Voice AI Models to Cresta's Contact Center Platform

Key Facts

  • Kotoba Technologies plans to partner with Cresta and provide its speech recognition and speech synthesis models to Cresta's Voice AI agents for contact centers.
  • Kotoba's Koto voice AI model is built for ultra-low-latency real-time processing optimized for East Asian languages.
  • The models support Japanese, Korean, Chinese, English, and other languages.
  • Provision begins sequentially in 2026 through Cresta's platform for its enterprise customers.
  • Kotoba is headquartered in San Francisco with its main R&D base in Chiyoda-ku, Tokyo; CEO is Noriyuki Kojima, and Cresta's CEO is Ping Wu.

Kotoba's model Koto is built for ultra-low-latency real-time processing tuned to East Asian languages. Provision starts sequentially in 2026, covering Japanese, Korean, Chinese, English and others, delivered through Cresta's platform to enterprise customers.

Cresta's contact center agents will get speech recognition and synthesis from Kotoba, per a partnership the two companies plan. Kotoba's Koto model is tuned for real-time processing in East Asian languages, and the company names Japanese as one of the hard cases: honorific expressions and a large number of homophones complicate recognition and synthesis.

Under the arrangement, STT converts customer speech to text for agent responses, operator support, and conversation analysis, while TTS reads agent replies with natural intonation and pauses. Provision begins sequentially in 2026 and reaches Cresta's enterprise customers through its platform.

Written with AI by Yomimono from the 1 cited source below, including Japanese-language reporting where cited. See our editorial policy.

Add Yomimono as a preferred source on Google

Sources