Airgentic Introduces Natural Voice Mode for Human-Like Customer Support

We are excited to share one of our biggest milestones yet. Airgentic voice mode is now live, bringing natural spoken interaction directly into our customer service platform. This release moves us closer to a future where a website can speak with customers as fluidly as a human contact centre assistant.

Unlike traditional voice interfaces that convert speech to text before processing, Airgentic streams audio directly into the language model. This reduces transcription errors and enables conversations that feel far more natural. Customers can interrupt, ask follow up questions and be heard immediately, just as they would in a real conversation.

Voice mode is fully integrated into Airgentic’s agentic workflow, making voice simply another powerful pathway for task specialist agents to resolve issues from start to finish. It also works seamlessly with our AI vision capabilities. This means customers can be guided through complex procedures from PDF manuals while the agent explains diagrams, labels and tables in real time.

Some key features include a spectrum analyser that visualises audio from both the customer and the AI, step by step guidance for long or detailed procedures, and in browser voice activity detection for reliable responsiveness. Voice mode supports multiple languages with configurable accents and offers flexible voice prompts. It is also tied into the Airgentic insights dashboard so teams can listen to conversations exactly as they occurred. All of this still runs entirely on Azure infrastructure.

Voice mode is currently in limited release as we monitor performance and feedback, with broader availability on the way. You can try it today on this website.

This update continues our commitment to delivering intelligent, grounded and effortless customer support experiences for every organisation we serve.