About the role
Deepgram is hiring a Director of AI/ML Research to lead its text-to-speech program, from research strategy and technical direction through models deployed in production. The role combines hands-on work on architectures, experiments, training strategy, and evaluation with responsibility for building the team and operating model.
The director will advance speech generation across prosody, expressiveness, controllability, multilingual speech, voice consistency, and inference performance, while measuring gains through automated metrics and human assessment. They will set research priorities under uncertainty, develop technical leaders, and partner with product and engineering on release readiness. Deepgram seeks deep expertise in modern TTS or audio generative modeling and a record of training and improving large-scale neural models. Experience scaling a research organization or deploying generative audio at production scale is a plus.
Highlights
- Own the TTS research and model roadmap from technical bets to production.
- Lead researchers and research engineers through technical leaders.
- Build evaluations that combine automated metrics and human perceptual assessment.
- Advance naturalness, expressiveness, controllability, multilingual output, and inference performance.
This recap is dataskew's editorial summary, not the company's copy.