Find your next role
Discover roles across ambitious teams
17,006 jobs
- More jobs
Deepgram
Research Staff Voice AI Foundations
Remote · USA · Senior level
$150k–$250k / year
About the role
Pioneer development of Latent Space Models to solve data, scale, and cost challenges in voice AI. Focus on neural audio codecs, generative models, embedding systems, synthetic data generation, and efficient model architectures.
What you’ll do
• Build next-gen neural audio codecs with low bit-rate compression. • Develop steerable generative models for diverse human speech synthesis. • Create embedding systems to factorize codec latent space. • Generate synthetic audio data at large scale via latent recombination. • Design model architectures and training for cost-efficient real-time inference.
What you’ll bring
• Strong foundation in statistical learning theory. • Expertise in foundation model architectures. • Ability to bridge theory and practical implementation. • Experience building scalable data pipelines. • Track record of experimental design and model optimization.
Life at Deepgram
AI-first mindset with rapid innovation and continuous learning. Emphasis on creative use of AI tools and fast-paced adaptation.
About Deepgram
Deepgram provides real-time APIs for speech-to-text, text-to-speech, and voice agents. It powers voice AI for 1,300+ organizations with unmatched accuracy and efficiency.