San Francisco-based Rime has raised $24 million in a Series A funding round led by M13 Ventures to advance its voice AI models for enterprise calls, as the startup tries to carve out an edge in a crowded market by training on proprietary conversational data and focusing on pronunciation accuracy. The investment signals that investors see room for voice AI to eventually match the reliability of legacy IVR systems in customer support.
Founded in 2022 by former Stanford PhD student Lily Clifford, ex-Amazon Alexa engineer Brooke Larson, and Stanford engineer Ares Geovanos, Rime operates a recording studio in San Francisco to collect its own conversational data instead of scraping the web. The company says this custom dataset lets it fine-tune models to handle brand names and industry-specific terms with high accuracy, reducing the customization load for clients. For customer support leaders tracking voice AI developments, resources like AI for Customer Support can help teams stay informed on the technology's evolving capabilities.
The stubborn gap between voice AI and IVR
Clifford said that despite progress in voice AI, enterprises still prefer legacy IVR implementations because the technology can't yet match IVR's effectiveness. "The voice technology is still not there to automate the vast majority of enterprise phone calls. LLMs have made it a lot easier to build voice applications that work, but they haven't changed how it feels to interact. Talking with a voice AI agent is not the most compelling experience for the end user. It's kinda like a new IVR, but with a better voice," she said.
Phoneme-based architecture and custom data
Rime employs a phoneme-based architecture to adapt to different pronunciations, so customers don't have to retrain models for their specific industry. The startup's recording studio captures conversational data that mirrors real-world enterprise calls, allowing it to tune voice models for precise pronunciation of brand entities and jargon. This approach aims to make interactions sound more natural without requiring heavy client-side customization.
Shifting from separate models to speech-to-speech
The company originally built a pipeline of separate models for speech-to-text, text-to-speech, and a large language model. It is now shifting to develop better speech-to-speech models to reduce latency, improve turn-taking, and tackle issues like background noise. The new approach also decreases reliance on orchestration so the company doesn't have to manage multiple models, potentially leading to smoother call handling.
Enterprise traction and expansion plans
Rime says it has customers in food service, healthcare, airlines, and fintech, including Mayo Clinic, Dialpad, Upstart, and Asurion. The company claims that its training data and model positioning keep callers on the line longer, helping it win enterprise contracts. With the new funding, Rime plans to expand its 35-person team, hiring for model development, engineering, and partnerships. It recently hired Rafael Valle, formerly of Meta Superintelligence Labs and Nvidia's applied deep learning audio research team, as chief scientist.
M13's Morgan Blumberg said, "Companies like ElevenLabs have moved into being an orchestration and the application layer, going head to head with the Sierras and Decagons of the world. I think there's just so much more to be done technically, and Rime's approach of pushing forward on the best model with low latency and high reliability in a regulated environment stands out." Blumberg is joining the startup's board as part of the fundraise.
Why this matters for customer support
As tools like Rime push voice AI closer to replacing parts of the call center, support leaders must weigh whether these models can reduce costs without harming the customer experience. The technology still can't fully replace IVR in high-stakes calls, but advances in speech-to-speech modeling and pronunciation accuracy could make AI agents more acceptable for handling sensitive interactions. For call center supervisors, a structured AI Learning Path for Call Center Supervisors can provide a framework for evaluating and adopting these tools.
Your membership also unlocks:
