Rasa, a leader in conversational AI, has introduced a groundbreaking advancement in enterprise voice automation. Showcased at Customer Contact Week (CCW) 2025 at Caesars Forum in Las Vegas, this next-generation voice architecture fuses structured logic with multimodal input to close the gap between real-time understanding and action — redefining how businesses engage with customers through voice.
Unlike traditional voice systems that rely on speech-to-text conversion, Rasa’s new voice automation technology processes spoken input directly. This eliminates common issues like lag, inaccuracies, and interruptions. With Rasa Voice, conversations are interpreted in real time, enabling systems to respond proactively — often before the speaker has even finished their sentence — resulting in faster, smoother, and more intuitive customer interactions.
“We believe that voice is becoming one of the most strategic channels for customer experience, and companies need reliable solutions,” says Melissa Gordon, Managing Director of Rasa. “What we’re introducing today is structured voice automation, built for scalability and capable of serving businesses with the speed, security, and nuance expected by the people using it.”
Enabling Conversational Journeys That Drive Outcomes
Rasa Voice is built to turn voice interactions into outcomes. It leverages live audio signals and contextual data to interpret user intent from the very first word — no transcription necessary. This enables:
-
Rapid execution without transcription delays
-
Enhanced recognition of intent, emotion, and language patterns
-
Natural, human-like interactions that maintain the integrity of spoken input
-
Seamless alignment between what is said and what the system delivers
Also Read: Movate & Alcatel-Lucent Enterprise Partner to Boost AI-Driven Digital Transformation
A Scalable, Reliable Solution Built on CALM Framework
At the core of Rasa Voice is the company’s proprietary CALM (Conversational AI with Language Models) framework. CALM marries the fluency of large language models with the dependability enterprises require. Instead of relying solely on prompting, CALM translates speech into clear, structured commands tied directly to business processes.
The result is consistent, high-performing voice automation that functions effectively across channels, under pressure, and at scale. It’s designed not just to sound natural, but to drive measurable business outcomes.
Optimized for Real-World Demands
Success in enterprise voice AI depends on speed, precision, and cost-efficiency. Rasa’s architecture is purpose-built with these priorities in mind. The platform allows businesses to select and fine-tune language models for their specific domains while meeting strict latency and throughput goals — minimizing risk and ensuring dependable performance in high-stakes environments.
“We don’t experiment with voice. We use it. That’s the difference,” says Loïc Mayet, Director of Information Systems at Groupe IMA. Groupe IMA, one of Europe’s top insurance and assistance providers, selected Rasa Voice after evaluating multiple platforms. “Rasa helped us develop the solution with our automation goals in mind and has been a partner every step of the way.”
Swisscom, Switzerland’s premier telecom provider, has also chosen Rasa Voice to expand its customer service automation without compromising on quality — further highlighting Rasa’s growing footprint among global enterprises.
