Machine Learning Scientist
Rime Labs · Remote · United States ·
COMPENSATION
€•• – ••k
Listed on the employer's job board (Ashby)
- Fully remote
- Visa sponsorship
- Speech synthesis focus
- PyTorch required
- PhD or equivalent
Rime develops voice AI infrastructure for enterprise customer experiences, focusing on high-volume conversational deployments. The team prioritizes data quality, utilizing a corpus recorded and annotated by PhD linguists to drive model performance.
As a Machine Learning Scientist, you will push the boundaries of speech synthesis and understanding. You will own the design, training, and evaluation of models while collaborating closely with linguists to refine frontend behavior and prosodic control.
What you'll do
- Design, train, and evaluate autoregressive and non-autoregressive speech synthesis models
- Research full-duplex and half-duplex multi-modal architectures including unified S2S systems
- Iterate on speech representations such as neural codecs, semantic tokens, and mel features
- Build rigorous objective and perceptual evaluation frameworks to ensure quality
- Collaborate with linguists to align TTS frontend behavior with modeling choices
What you'll need
- Deep knowledge of speech synthesis literature including Tacotron, FastSpeech, VITS, and VALL-E
- Hands-on experience with neural codecs like EnCodec, DAC, or Mimi
- Experience with full- or half-duplex multi-modal modeling such as Moshi or LLaMA-Omni
- Strong attention to data quality and ability to detect annotation pipeline issues
- Working knowledge of TTS frontend components like G2P, normalization, and prosody
- Strong PyTorch fundamentals including distributed training and model internals
- PhD or equivalent research experience in speech, audio, ML, or computational linguistics
Nice to have
- Experience with multilingual TTS systems
- Background in prosody or paralinguistics
- Published work in speech, audio, or core ML venues
- Experience taking research models to production via quantization or distillation
Stack
- PyTorch
- EnCodec
- DAC
- Mimi
- Tacotron
- FastSpeech
- VITS
- VALL-E
How you'll work
- Fully remote policy
- Visa sponsorship available
What you get
- Competitive base salary
- Meaningful early-stage equity
- Access to proprietary studio-quality speech corpus
- High ownership and low bureaucracy culture
- Direct collaboration with founders