Cartesia is looking for a Researcher: Audio to join our UK team in London. You will conduct pioneering research at the intersection of audio signal processing, machine learning, and generative modeling to push the boundaries of voice AI systems.

What You'll Do

Conduct research at the intersection of audio signal processing, machine learning, and generative modeling to advance voice AI.
Develop cutting-edge algorithms for tasks like speech enhancement, echo cancellation, denoising, and voice activity detection using generative approaches.
Design novel methods for end-to-end modeling of audio signals, exploring neural audio synthesis and self-supervised training.
Lead the development of robust evaluation pipelines to analyze performance and identify future research directions.

What We're Looking For

Deep expertise in audio signal processing, generative modeling, and machine learning.
Proven track record of publishing impactful research in top-tier conferences like NeurIPS, ICASSP, or ICLR.
Proficiency in frameworks such as PyTorch or TensorFlow and specialized audio tools like torchaudio or librosa.
Strong understanding of state-of-the-art generative techniques, including diffusion models, autoregressive models, and flow-based models.
Passion for solving complex problems in speech and audio, with a focus on creating innovative, practical solutions for noisy, multi-modal, or real-time environments.
Excellent collaboration and communication skills, with the ability to work effectively in research-driven and cross-functional teams.

Nice to Have

Experience building audio models that have been used in production at scale.
Background in hardware-aware optimization for deploying real-time audio models.
Early-stage startup experience or experience working in fast-paced R&D environments.