Rime Launches the Prism Speech-to-Speech Model 

Rime, a developer of enterprise voice artificial intelligence models, today launched Prism, a customizable frontier speech-to-speech model that companies can run entirely within their own infrastructure while maintaining control over their data, conversation latency, vocabulary, pronunciation, and safety guardrails.

Prism has been trained on millions of hours of human-generated conversations captured in Rime's sound studios. Key features include the following:

• Architectural Guardrails: Generates text before speech, allowing internal rules and policies to review every word before the customer hears it.
• Total Deployment Ownership: Run the model on-premise or in private cloud, while remaining fully SOC 2 and HIPAA compliant.
• Flawless Auditability: Since PRISM generates text first, businesses have quick and complete audit trails.
• Human-Grade Emotional Understanding: Trained to understand customer emotions to better support their requests and route their calls.
• Advanced Conversational Intelligence: Built on proprietary RL post-training using thousands of hours of human conversations, PRISM catches customer frustration, hesitations, acting instantly to eliminate dead air and improve containment rates.

"A generic model with a custom prompt is not your voice. It's a rental with your logo on it," said Lily Clifford, co-founder and CEO of Rime, in a statement. "Prism is the first speech-to-speech model that becomes yours. It learns from your conversations, lets you set the latency and guardrails, and lives on your infrastructure. Quickly, the model running your calls will be one no competitor can buy."