rime raises $24m for voice ai trained on its own data

source: techcrunch ai: rime picks up $24m series a to help enterprises field customer calls

level: business

rime, a san francisco voice ai startup, raised $24 million in series a funding led by m13 ventures, with participation from twilio ventures and others. founded in 2022, rime builds voice models trained on data from its own recording studio, rather than scraping the web. this approach helps the models pronounce brand names and industry terms accurately using a phoneme-based architecture, reducing the need for client customization.

the company initially used separate models for speech-to-text, text-to-speech, and language processing, but is now shifting to speech-to-speech models to cut latency and improve turn-taking. rime targets enterprises in food service, healthcare, airlines, and fintech, claiming its models keep callers engaged longer. clients include mayo clinic, dialpad, and upstart. the funding will expand its 35-person team, with a focus on model development and partnerships, and it recently hired a chief scientist from meta and nvidia.

despite advances in voice ai, many enterprises still rely on traditional interactive voice response systems because ai voice interactions often feel unnatural. rime aims to change that by focusing on low-latency, reliable models suited for regulated environments. the startup competes with companies like elevenlabs and sierra, but its data collection strategy and technical focus on core model performance set it apart, according to investors.

why it matters: better voice ai can automate more enterprise calls, reducing costs and improving customer experience in industries like healthcare and finance.

source: techcrunch ai: rime picks up $24m series a to help enterprises field customer calls