PhD in privacy-preserving speech. Co-creator of the VoicePrivacy Challenge.

I am the co-founder and CEO of Nijta, a deep-tech startup in Lille that makes voice anonymization usable at scale. We turn recorded speech into privacy-safe audio by removing the biometric voiceprint and personal information while keeping the content usable.

Before Nijta, I completed my PhD at Inria (Magnet and Multispeech teams) on privacy-preserving speech, supervised by Aurélien Bellet, Emmanuel Vincent and Marc Tommasi, and I worked as a research engineer at Microsoft Research on multilingual speech recognition. I co-created the VoicePrivacy Challenge, the community benchmark for voice anonymization, and I serve on the organizing committee of the SPSC Symposium.

My work sits at the intersection of speech, machine learning and privacy. If you would like to know more about Nijta or the VoicePrivacy Challenge, feel free to reach out.

Google Scholar, as of July 2026: about 1,500 citations, h-index 15, i10-index 22.