Metavoice is a London-based voice AI startup building next-generation speech-to-speech technology designed to make voice the most natural way to interact with AI. Backed by investors including 7percent Ventures and Entrepreneur First, their research focuses on duplex conversational models that are emotionally aware and naturally responsive.

I worked with MetaVoice as a Recording Engineer and Procution Lead, collaborating directly with the co-founder to optimise their voice data capture pipeline. My role involved designing and implementing a scalable recording workflow for high-quality speech data collection, ensuring consistent capture standards across sessions to produce the precise data points needed to improve their machine learning models.