Hello!

I am the Head of Machine Learning Research at Sonos, focusing on privacy-first speech recognition for voice experiences. I am lucky to work with a team of around 60 extremely talented and passionate people in our R&D department “Sonos Voice Experience” in Paris.

I own the scientific aspect of the R&D for Sonos Voice Control. My main mission is to promote a fruitful and ambitious scientific research environment in the machine learning team that boosts and supports our production goals: building a fully private voice control solution on Sonos speakers.

I am a machine learning project manager for the next generations of our product and a technical project lead in several key areas (e.g. fairness, automatic speech recognition). I also have had and continue having individual contributions on a variety of topics in the field, such as natural language processing, keyword spotting, speaker identification, directedness classification, and more recently, fairness.

On June 1st 2022, after several yearks of work, we released Sonos Voice Control: an on-device voice control interface on Sonos speakers, putting speed accuracy and privacy on an equal footing. You can find more details in this post on Sonos Tech Blog.

I hold a PhD in statistical physics from the École Normale Supérieure (Paris) since 2016. I focused on high-dimensional inference with graphical models in the context of protein structure prediction.

Contact: coucke [at] phare [dot] normalesup [dot] org

LinkedIn: acoucke

Twitter: @alicecoucke