Andy Jassy, CEO, Amazon
Amazon’ s Nova Sonic is a unified voice AI model that can make virtual assistant conversations feel human.
The virtual assistant tackles the problem of traditional voice AI systems sounding emotionally flat and robotic because they strip away crucial conversational nuances.
Danilo Poccia, Chief Evangelist( EMEA) at AWS, explains that conventional approaches require“ complex orchestration of multiple models, such as speech recognition to convert speech to text, language models to understand and generate responses, and text-tospeech to convert text back to audio.”
This fragmented system“ increases development complexity but also fails to preserve crucial linguistic context such as tone, prosody and speaking style.”
What makes Nova Sonic unique is that it integrates speech understanding and generation into one model.
It intelligently manages pauses, interruptions and hesitations while adapting responses based on users’ speaking styles – preserving the emotional context that makes conversations feel natural.
Available through Amazon Bedrock’ s API, the model targets industries from healthcare to customer service where nuanced interaction matters.
218 October 2025