ML Engineer, Speech & Voice Synthesis

Cantina Labs; United States; Remote; USD 180,000 - 260,000; Full time

Benefits offered by this job

Competitive salary & equity

Medical, dental, and vision insurance

42 days of paid time off (PTO, sick, &

Parental leave & fertility support

401(k) retirement savings plan

Lifestyle spending account – $500/mo

Complimentary lunch and snacks

One Medical membership

Job summary

Cantina Labs is seeking a Research / ML Engineer to join the Speech Team. You will build state-of-the-art speech systems end-to-end, from data specs to production inference, and drive the model data evaluation flywheel for controllable TTS and voice design.

You will collaborate with research, data, and infra to ship reliable, cost-aware models while shaping safe and steerable AI systems.

Qualifications

• Extensive experience with large-scale audio models and data.
• Hands-on work with diffusion/flow-matching transformers and samplers.
• Experience training audio VAEs, codecs, and vocoders.
• Proficiency in multi-node, multi-GPU distributed training.
• Strong software engineering and production-quality code.
• Expertise in PyTorch; CUDA/Triton/C++ performance work.
• Shipped speech/audio or multimodal models to production.
• Experience with large-scale ML data and evaluation signals.
• Experience with voice cloning or expressive speech generation.
• Notable publications or open-source contributions.

Responsibilities

• Model building: pre-train, fine-tune, and align large-scale speech models.
• Design experiments and analyze results to improve models.
• Develop dev tooling to boost team productivity.
• Contribute to full stack from low-level optimizations to model design.
• Define data requirements and guide data curation and labeling.
• Design automated evaluations including listening tests and ASR metrics.
• Harden training to inference pipelines; monitor latency and cost.
• Contribute to safety guardrails and responsible AI practices.

Skills

Large-scale audio models

Diffusion models

Audio VAEs & Vocoders

Distributed training

Software engineering

PyTorch & CUDA

Production models

ML data expertise

Voice cloning

Publications / OSS

Tools

FSDP/DeepSpeed

Job description

Cantina Labs is seeking a Research / ML Engineer to join the Speech Team. You will build state-of-the-art speech systems end-to-end, from data specs to production inference, and drive the model data evaluation flywheel for controllable TTS and voice design.

You will collaborate with research, data, and infra to ship reliable, cost-aware models while shaping safe and steerable AI systems.