Senior Research Scientist
Core
Post-training large speech models to handle natural, messy human conversations by curating data, fine-tuning, and building evaluation infrastructure.
Role type
Senior Research Scientist (Speech AI)
Builds
Large-scale speech models for natural conversation
Domain
Voice AI / Generative AI
Deliverable
production ML models
Required skills
PhD in ML, training large speech models (ASR/TTS/speech-to-speech), post-training and SFT, data curation, building evaluation workflows
Preferred skills
strong publications record, experience with hallucination reduction and instruction-following optimization
Technologies
ASR, TTS, speech-to-speech
Responsibilities
Source, clean, and structure data for large speech models; perform supervised fine-tuning of pre-trained models; build automated and human-in-the-loop evaluation workflows; drive improvements in hallucination rates, instruction-following, and generalisation
Seniority
Senior, hands-on IC