Rime secures $24 million Series A to refine voice AI for enterprise customer service
Led by M13 Ventures, the funding round supports Rime’s shift from legacy IVR alternatives to proprietary conversational data strategies for healthcare, fintech, and airline sectors.
San Francisco-based voice artificial intelligence startup Rime has secured $24 million in Series A funding, led by M13 Ventures. The investment round also included participation from Twilio Ventures, Corazon Capital, and Unusual Ventures, with existing investors joining the syndicate. As part of the agreement, Morgan Blumberg from M13 Ventures will join Rime’s board of directors.
Founded in 2022 by Lily Clifford, Brooke Larson, and Ares Geovanos, Rime specialises in voice models trained on proprietary conversational data. Unlike competitors that rely on web-scraped audio, the company operates a dedicated recording studio in San Francisco to collect its own data. This strategy aims to improve the pronunciation of industry-specific terms and brand entities, reducing the customization load for enterprise clients across sectors such as healthcare, airlines, fintech, and food service.
The startup currently processes over 100 million calls monthly for clients including Mayo Clinic, Dialpad, Upstart, and Asurion. Rime employs a phoneme-based architecture designed to adapt to different pronunciations without requiring clients to retrain models for their specific industries. The company claims that its training data and model positioning help customers stay on calls longer, a metric that has assisted in securing these enterprise contracts.
Technically, Rime is shifting its focus from separate pipelines for speech-to-text, text-to-speech, and large language models to integrated speech-to-speech models. This transition aims to reduce latency, improve turn-taking, and mitigate issues such as background noise. The new approach also seeks to decrease reliance on complex orchestration, allowing the company to manage a single model rather than multiple components.
Despite advancements in voice AI, Rime co-founder Lily Clifford noted that many large organisations still prefer legacy Interactive Voice Response (IVR) systems due to reliability concerns. Clifford stated that while large language models have made building voice applications easier, the user experience has not significantly improved, describing current AI voice agents as similar to IVR but with better voice quality.
To support its technical evolution, Rime recently appointed Rafael Valle, formerly of Meta Superintelligence Labs and NVIDIA, as its Chief Scientist. The company, which employs 35 people, plans to use the new capital to expand its team in model development, engineering, and partnerships. This Series A follows a $5.5 million seed round raised in May of the previous year.
The enterprise voice AI market remains highly competitive, with established players such as ElevenLabs, Deepgram, Vapi, Retell, LiveKit, Decagon, and Sierra vying for market share. M13 Ventures’ Morgan Blumberg highlighted Rime’s distinct approach, noting that while some competitors have moved into orchestration and application layers, Rime is focusing on delivering low-latency, high-reliability models suitable for regulated environments.
