Director, Text-to-Speech Synthesis Research

About the role

The Director of Research will own the end-to-end Text-to-Speech program, including research strategy, technical bets, and model deployment. This is a hands-on leadership role where you will shape architectures, experiments, and training strategies while building and managing a team of researchers. The mandate is to advance the state of the art in speech generation and ensure these advances translate into measurable production gains.

Responsibilities

• Own the TTS research and model roadmap, deciding which technical directions to pursue and when to pivot.
• Drive advances in neural audio modeling, prosody, expressiveness, and inference performance.
• Stay deeply technical by reviewing research, designing experiments, and diagnosing model failures.
• Build evaluation and benchmarking systems that combine automated metrics with human perceptual assessment.
• Lead a team of individual contributors and tech lead managers, hiring and developing talent while maintaining a high technical bar.
• Partner with engineering and product leadership to ensure research translates effectively into production.

Required skills

• Text-to-Speech
• Speech generation
• Audio generative modeling
• Neural models
• Research leadership
• AI

Nice to have

• Production-scale deployment
• Research organization scaling
• Evaluation systems
• Multilingual generation
• Voice cloning
• Controllable generation

About the Company

Deepgram is a leading platform in the Voice AI economy, providing real-time APIs for speech-to-text and text-to-speech. The company supports over 200,000 developers and 1,300 organizations with voice-native foundation models that offer high accuracy and low latency.