Rime raises $24M to put AI voices on enterprise phone lines
The voice on the other end of your next customer-service call is increasingly not a person, and Rime just raised $24 million in a Series A to make more of those voices its own. The round was led by M13, with Twilio Ventures, Corazon Capital and returning backer Unusual Ventures joining in, and M13 partner Morgan Blumberg taking a board seat.
Rime, run by CEO Lily Clifford, builds what it calls the first enterprise-ready speech-to-speech model, taking a linguistics-first approach rather than bolting a voice onto a text model after the fact. The traction is the headline: it already powers close to 100 million phone calls a month for names like Mayo Clinic, Dialpad, Upstart and Asurion. That's not a demo, that's production voice AI handling real customers at real scale.
Speech-to-speech matters for agents specifically because the old pipeline, transcribe to text, run an LLM, synthesize speech, adds latency and strips out everything about how someone actually said something. A native speech model can hear hesitation and interruption and respond in the rhythm of a real conversation, which is the difference between an agent that feels like a phone tree and one that feels like a person.
The fresh capital goes into Rime's proprietary conversational dataset and engineering hires, which is the honest tell: in voice AI the moat is the data, not the demo. With 100 million monthly calls flowing through, Rime is compounding exactly the asset that's hardest to buy. Enterprise phone lines are quietly becoming one of the largest live deployments of AI agents anywhere, and most people will never know they were talking to one.
More at https://rime.ai/
← Back to all articles
Rime, run by CEO Lily Clifford, builds what it calls the first enterprise-ready speech-to-speech model, taking a linguistics-first approach rather than bolting a voice onto a text model after the fact. The traction is the headline: it already powers close to 100 million phone calls a month for names like Mayo Clinic, Dialpad, Upstart and Asurion. That's not a demo, that's production voice AI handling real customers at real scale.
Speech-to-speech matters for agents specifically because the old pipeline, transcribe to text, run an LLM, synthesize speech, adds latency and strips out everything about how someone actually said something. A native speech model can hear hesitation and interruption and respond in the rhythm of a real conversation, which is the difference between an agent that feels like a phone tree and one that feels like a person.
The fresh capital goes into Rime's proprietary conversational dataset and engineering hires, which is the honest tell: in voice AI the moat is the data, not the demo. With 100 million monthly calls flowing through, Rime is compounding exactly the asset that's hardest to buy. Enterprise phone lines are quietly becoming one of the largest live deployments of AI agents anywhere, and most people will never know they were talking to one.
More at https://rime.ai/
Comments