Job Title: Speech ML Research Contractor
Location: Remote (UK)
Type: 3–6 Month Contract (Full-Time or Part-Time: 20 hrs/wk for PhD students)
Role Overview Join an advanced speech research team conducting exploratory, hypothesis-driven research in text-to-speech (TTS), automatic speech recognition (ASR), and generative audio models.
Open to current PhD candidates (flexible ~20 hrs/week) or recent PhD/MSc graduates (full-time).
What You'll Do
- Research and build models for TTS, ASR, voice conversion, and audio generation.
- Experiment with modern generative architectures: Transformers, Diffusion, Flow Matching, GANs, VAEs, and Neural Audio Codecs.
- Clean, analyze, and evaluate real-world audio datasets at scale.
- Collaborate with research scientists with opportunities to publish at top-tier ML venues.
Requirements
- Education: Currently enrolled in or recent graduate (within 1 yr) of a PhD or MSc in ML, Signal Processing, or CS.
- Core Stack: Hands-on experience with PyTorch and handling complex data pipelines.
- ML Depth: Strong theoretical grasp of generative modeling (Diffusion, Flow Matching, Codecs).
- Skills: Independent, curious, and comfortable with open-ended research problems.
Interview Process
- Technical Interview (60 min): ML depth and research breadth discussion (no live coding).
Values Interview (30 min): Collaboration, communication, and team alignment.