About the role
Cantina is hiring a Machine Learning Engineer for its Speech Team to build text-to-speech and related audio systems, including voice cloning, controllable TTS, and voice conversion. The role spans research and production engineering, from model and data specifications through training, evaluation, and inference. You would design experiments, develop large-scale speech models, and work with research, data, and infrastructure partners to ship reliable, cost-aware systems.
The work also includes defining data requirements, improving training and evaluation pipelines, scaling distributed workloads across GPUs, and measuring quality, latency, memory use, and cost. The team is seeking hands-on experience with large audio models, transformer or diffusion architectures, distributed training, PyTorch, and production speech systems. The listing specifies remote work in Europe.
Highlights
- Build and deploy large-scale speech models, including TTS and voice conversion.
- Design experiments and objective and subjective model evaluations.
- Own parts of data curation, training pipelines, and production inference.
- Contribute to speech safety, consent guardrails, and misuse mitigation.
This recap is dataskew's editorial summary, not the company's copy.