We are recruiting a fully funded, three-year PhD candidate for a project on Privacy-Preserving Speech Understanding with Multimodal Signals for Clinical Applications, hosted by the MULTISPEECH team at LORIA (Université de Lorraine, Inria and CNRS), in Villers-lès-Nancy, France.

The project is funded by the AI Grand Est ENACT research chair and will investigate methods to protect speaker identity and sensitive content while preserving the semantic and diagnostic information needed for clinical speech understanding. The work will combine speech and text, with the possibility of incorporating an additional modality such as physiological signals, medical imaging or electronic health-record metadata. It will involve privacy-preserving speech processing, speech/audio foundation models, multimodal machine learning, and evaluation of privacy–utility trade-offs.

We welcome candidates with a Master’s or engineering degree in computer science, AI, signal or speech processing, applied mathematics, data science, or a related area. Strong Python and machine-learning/deep-learning skills are expected. Experience in speech processing, NLP, privacy-preserving ML, multimodal learning, or AI for healthcare is particularly welcome.

• Location: Villers-lès-Nancy, France
• Duration: 3 years, 1 October 2026 to 30 September 2029
• Start date: 1 October 2026 (or 1 November 2026)
  
• Salary: €2,300 gross per month
• Application deadline: Applications reviewed until the position is filled; final deadline 30 August 2026, potentially extended to 20 September 2026

Applicants should email a CV, motivation letter, degree transcripts, and, if available, two recommendation letters or the contact details of two referees.

For the full position description and application details, please see: https://sites.google.com/view/natalia-tomashenko/recruitment-phd-positions Contact: Natalia Tomashenko, natalia.tomashenko@inria.fr