Director, Text-to-Speech Research & Production

Deepgram; United States; Remote; USD 200,000 - 360,000; Full time

Job summary

Deepgram is seeking a Director of Research to own our Text-to-Speech program end to end—from strategy to models that ship. This is a hands-on leadership role, shaping architectures, experiments, training strategy, and evaluation while building a team and operating model that enables fast, high-quality progress.

You will lead research across neural audio modeling, prosody, multilingual speech, and deployment, partnering with engineering and product to ensure ship-readiness and measurable

Qualifications

• Deep expertise in modern TTS, speech generation, or audio generative modeling.
• Command of modern speech-generation stack and open problems: naturalness, expressiveness, controllability, robustness, inference cost.
• Experience setting research direction under uncertainty: prioritizing experiments and allocating compute.
• Experience leading researchers and researchers engineers through other leaders, building tech-lead managers.
• AI as default mode of work with a strategic view of its limits in speech research.
• Ability to communicate complex tradeoffs to product, engineering, and execs.

Responsibilities

• Own the TTS research and model roadmap and decide on directions with high impact.
• Drive advances across neural audio modeling, prosody, multilingual speech, and production gains.
• Stay deeply technical: review research, design experiments, diagnose failures, tackle high-leverage problems.
• Build evaluation and benchmarking showing why models improve, with automated metrics and human assessment.
• Lead a mix of ICs and tech leads, hire and develop talent, set direction across sub-teams.
• Partner with engineering and product leadership on ship-readiness and represent Deepgram’s TTS research externally.

Skills

TTS expertise

Speech generation

Neural models

Lead researchers

AI-first mindset

Job description

Deepgram is seeking a Director of Research to own our Text-to-Speech program end to end—from strategy to models that ship. This is a hands-on leadership role, shaping architectures, experiments, training strategy, and evaluation while building a team and operating model that enables fast, high-quality progress.

You will lead research across neural audio modeling, prosody, multilingual speech, and deployment, partnering with engineering and product to ensure ship-readiness and measurable