Research Scientist - Generative Audio
Core
Conduct groundbreaking research in generative audio using diffusion or flow matching models to create breakthrough listening experiences for artists and fans.
Role type
Research Scientist (Generative Audio)
Builds
New generative music products and tools for Spotify's platform
Domain
Music technology, Generative AI, Audio processing
Deliverable
production ML models
Required skills
Generative modeling, Machine learning, Vocal synthesis, Post-training alignment, Audio editing, Python, PyTorch, NumPy, Signal processing
Preferred skills
Music information retrieval, Computer vision, Probabilistic modeling
Technologies
Diffusion models, Flow matching, DPO, RLHF, KTO, PPO, GRPO
Responsibilities
Run large-scale experiments with Spotify's infrastructure, Create practical applications for generative technologies, Collaborate with cross-functional teams, Publish findings and deliver talks at top conferences
Seniority
All levels (Ph.D. required)