Rime raises $24 million Series A to help organizations make customer calls

The biggest launch for AI startups has been call handling for enterprises in areas such as sales, marketing, and customer support. Large organizations offload calls to voice model developers such as Eleven laboratories and Deepgram; Infrastructure companies such as Coat of armsRetail and LifeKit; and dedicated customer support stores such as Decagon and Sierra.

Based in San Francisco Reem It is trying to get an edge in this crowded market with voice AI models that are trained on the conversation data it records, with the aim of reducing the burden of personalization on its customers.

Founded in 2022 by former Stanford PhD student Lily Clifford, former Amazon Alexa engineer Brooke Larson, and Stanford engineer Ares Geovanos, Rime built a recording studio in San Francisco to collect its own conversational data rather than rely on web scraping for audio.

The startup said it is focusing on fine-tuning its voice models to be able to pronounce different brand entities and industry-specific terms. It uses audio-based architecture to adapt to different pronunciations so customers don’t have to retrain models specific to their specific industry.

Rime said Wednesday it has raised $24 million in a Series A funding round led by M13 Ventures. Twilio Ventures, Corazon Capital, Unusual Ventures and other existing investors also participated.

Clifford said that despite progress in the development of voice AI, companies still prefer legacy IVR (interactive voice response) applications, as voice AI technology is still unable to match the effectiveness of IVR.

“Voice technology still doesn’t exist to automate the vast majority of enterprise phone calls. LLMs have made it much easier to build effective voice applications, but they haven’t changed how we interact. Talking with a voice AI agent isn’t the most compelling experience for the end user. It’s a bit like the new Interactive Voice Response (IVR) system, but with a better voice.”

The startup started with a set of separate models for speech-to-text, text-to-speech, and a large language model. But it is now shifting focus to developing better speech-to-speech models to reduce latency, improve turn-taking, and address issues such as background noise. The new approach will also reduce reliance on coordination, so that the company does not have to manage a range of forms.

Rime says it has clients in food service, healthcare, airlines and fintech. The company claims that because of its training data and model positioning, customers stay longer on a call, which has helped it win enterprise contracts from clients like Mayo Clinic, Dialpad, Upstart and Asurion.

With the new funding, Rime plans to expand its team of 35 people, with the aim of hiring them for model development, engineering and partnerships. Rafael Valle, who worked on understanding sound at Meta Superintelligence Labs and the Applied Deep Learning Audio Research team at Nvidia, was recently named chief scientist.

“Companies like ElevenLabs have moved on to being an orchestration and application layer, competing head-to-head with the Sierras and Decagons of the world. I think there’s a lot more to be done from a technical standpoint, and Rime’s approach stands out in moving toward the best model with low latency and high reliability in a regulated environment,” M13’s Morgan Blumberg told TechCrunch.

It had previously raised $5.5 million Seed round last May. Blumberg is joining the startup’s board of directors as part of the fundraising campaign.