New features bring faster, more natural interactions and deeper enterprise control

boost.ai announced a series of upgrades to its Conversational AI platform, delivering a next-generation voice experience for enterprises. The upgraded voice offering introduces Express Voice for instant, speech-to-speech conversations, and enhancements to Enterprise Voice for complex and regulated scenarios. boost.ai is also rolling out new features, including real-time transcription, new performance analytics, speech-to-text fine-tuning, and natural background audio that makes conversations feel more human.

According to Gartner, by 2028, 70% of customer service journeys will start and end with voice-based assistants on mobile, and by 2029, agentic AI will resolve 80% of issues autonomously. As enterprises face this shift toward voice-first engagement, they need solutions that can balance speed, natural flow, and enterprise-grade reliability. Boost.ai’s upgraded voice offering is designed to deliver on all three.

“Customers today expect voice interactions that are as seamless as a real conversation with a human,” said Jerry Haywood, CEO of boost.ai. “With our upgraded voice offering, we’re giving enterprises the ability to meet those expectations instantly, while still ensuring the control and reliability they need. Voice Agents are moving to the center of customer service, and those who get it right now will own the loyalty of tomorrow’s customers.”

Marketing Technology News: MarTech Interview with Stephen Howard-Sarin, MD of Retail Media, Americas @ Criteo

Boost.ai’s upgraded voice offering introduces two distinct modes of deployment designed to meet the diverse needs of modern enterprises. Express Voice, powered by speech-to-speech, enables instant, fluid conversations by sending audio directly to multimodal models without transcription delay. The result is instant, seamless and human-like responses, ideal for starting conversations and resolving simple, FAQ-style inquiries. Enterprise Voice, based on a traditional Speech-To-Text/Text-To-Speech setup, provides the full control and reliability that regulated industries require. With support for multi-intent and transactional conversations, built-in PII masking, hybrid NLU, and robust guardrails, it ensures every interaction is secure, predictable, and compliant.

The upgrade also introduces new tools that raise both quality and performance. Real-time transcription during test calls makes it simple to verify recognition accuracy and quickly correct misheard inputs. Voice analytics deliver essential call metrics from day one, giving enterprises clear visibility into performance at scale. With speech-to-text fine-tuning, organizations can add brand-specific vocabulary such as product names and acronyms to ensure accurate speech recognition. And with ambient and thinking audio, conversations feel smoother and more natural — replacing silences with subtle cues that keep customers engaged and confident that the AI is actively listening.

Marketing Technology News: From MarTech Stack to MarTech Fabric: Weaving Brand, Content, and Conversion Into One Thread

“Voice is fast becoming the default interface for customer service,” said Rasmus Hauch, CTO of boost.ai. “By offering both instant, natural interactions and enterprise-grade control, we’re giving businesses the tools to interact with their customers in the most natural way possible, without compromising on safety, accuracy, or trust.”

The next evolution of boost.ai’s voice offering will be an industry-first hybrid approach, Adaptive Voice, which dynamically blends Express and Enterprise capabilities in a single conversation. It enables fast, free-flowing interactions for simple use cases, while seamlessly switching to full compliance and control when complexity or regulation requires it. A future-ready approach that delivers both agility and assurance, without compromise.

The upgraded voice offering is available today for boost.ai customers, underscoring the company’s commitment to evolving the customer experience and helping enterprises succeed in a voice-first world.