Research Engineer, Audio

As a Research Engineer at ElevenLabs, you will be at the forefront of the generative AI revolution, building the most advanced text-to-speech and audio foundational models in the world. You will bridge the gap between academic paper concepts and production-ready architectures, optimizing large-scale models that can synthesize human-quality speech and sound effects in real-time. Join our globally distributed team to redefine how the world interacts with digital content through emotionally rich, low-latency audio synthesis.

- Design and implement novel architectures for multi-lingual text-to-speech and expressive voice cloning. - Scale training pipelines for foundational audio models using distributed training techniques and performance optimization. - Develop data curation and preprocessing workflows for massive-scale, diverse audio datasets. - Rapidly prototype and iterate on research ideas from recent literature to improve audio fidelity and latency. - Collaborate with infrastructure engineers to deploy models into a high-throughput production environment. - Optimize inference engines for low-latency delivery across various hardware backends. - Monitor and evaluate model performance using both objective metrics and subjective human-in-the-loop testing.

- Master’s or PhD in Computer Science, Machine Learning, or a related quantitative field. - Proficiency in Python and deep learning frameworks, specifically PyTorch and JAX. - Strong mathematical foundation in probability, statistics, and signal processing. - Proven track record of training large-scale generative models (GANs, Diffusion, Transformers). - Experience with high-performance computing clusters and distributed training across hundreds of GPUs. - Deep understanding of modern audio synthesis techniques such as Vocoding, WavNet, or Flow-based models. - Excellent communication skills for a fully remote, asynchronous work environment.