Director of TTS Research: Lead Frontier Speech Models

Deepgram; Greater London; On-site; GBP 150,000 - 210,000; Full time

Job summary

Deepgram seeks a Director of Research to own the Text-to-Speech program end to end, from strategy to deployment. You will guide architectures, experiments, training strategies, and evaluation while building a team of researchers and tech leads to move fast without lowering the bar.

You will collaborate with engineering and product leadership to ensure ship-readiness, and you’ll represent the TTS research program both internally and externally as you push for breakthrough, production-grade voice

Qualifications

• Deep expertise in modern TTS, speech generation, or audio generative modeling.
• Track record of training and improving large-scale neural models.
• Experience leading researchers and research engineers across sub-teams.
• Experience in fast-moving startup or research environments bringing models to production.

Responsibilities

• Own the TTS research roadmap and model strategy across architectures and training.
• Lead a mix of ICs and tech lead managers while maintaining high technical bar.
• Build evaluation systems with automated metrics and perceptual assessment.
• Partner with engineering and product leadership on ship-readiness.
• Represent the company’s TTS research internally and externally.

Skills

TTS expertise

Speech generation

Audio generative modeling

Team leadership

Experimental design

Research direction

Production-scale models

Communication with leadership

Job description

Deepgram seeks a Director of Research to own the Text-to-Speech program end to end, from strategy to deployment. You will guide architectures, experiments, training strategies, and evaluation while building a team of researchers and tech leads to move fast without lowering the bar.

You will collaborate with engineering and product leadership to ensure ship-readiness, and you’ll represent the TTS research program both internally and externally as you push for breakthrough, production-grade voice