Research Staff – Voice AI Foundations

Deepgram

. Develop next-generation neural audio codecs with extreme low-bit-rate compression and high-fidelity reconstruction .

Posted 10/11/2026full-timeRemote • AustraliaLeadWebsite

Core Competencies

Role fit

Core Competencies

Use this summary to align your resume positioning with the role.

Demonstrates expertise in developing neural audio codecs and generative models, with a strong foundation in statistical learning theory and experience in optimizing models for real-world deployment. Proven ability to design experiments and curate large datasets while advancing research in voice AI and multimodal learning.

Highest-signal resume keywords

Neural Audio Codec DevelopmentGenerative Model ArchitectureStatistical Learning TheoryMultimodal LearningReal-World Model Optimization

ATS Keywords

Tailor your resume

Applicant Tracking System Keywords

Tip: use these terms in your resume and cover letter to boost ATS matches.

Hard Skills

Neural Audio CodecsGenerative ModelsStatistical Learning TheoryMultimodal Speech-to-Speech SystemsModel Architecture DesignTraining Scheme DevelopmentInference Algorithm OptimizationDataset CurationMathematical FormulationControlled Experiment Design

Soft Skills

Critical ThinkingProblem SolvingCollaboration

Tools & Technologies

AI ToolsBare-Metal Hardware

Industry Keywords

Voice AISpeech ModelingLanguage ModelingLatent RepresentationsEmbedding Systems

About the role

Key responsibilities & impact

Develop next-generation neural audio codecs with extreme low-bit-rate compression and high-fidelity reconstruction
Pioneer steerable generative models for diverse human speech, including emotional, multi-speaker, noisy, and overlapping-speech scenarios
Develop embedding systems that factorize latent representations into speaker, content, style, environment, and channel dimensions
Use latent recombination to generate synthetic audio data at previously impossible scales
Train multimodal speech-to-speech systems for robust understanding and empathic human-like responses
Design model architectures, training schemes, and inference algorithms optimized for bare-metal hardware
Enable cost-efficient training on billion-hour datasets and real-time inference for hundreds of millions of concurrent conversations
Conduct rigorous controlled experiments, ablations, evaluations, stress tests, and edge-case analyses
Contribute to foundational research advancing voice AI, speech, and language modeling

Requirements

What you’ll need

Strong mathematical foundation in statistical learning theory, particularly self-supervised and multimodal learning
Deep expertise in foundation model architectures and scaling training across multiple modalities
Ability to derive novel mathematical formulations and implement them efficiently
Demonstrated ability to build and curate massive datasets while maintaining quality and diversity
Track record of designing controlled experiments to validate architectural innovations and theoretical insights
Experience optimizing models for real-world deployment, including hardware constraints and efficiency techniques
History of open-source contributions or research publications advancing speech/language AI
Comfort actively using and experimenting with advanced AI tools
Ability to identify critical experiments that validate or disprove ideas quickly
Vision to scale successful proofs-of-concept 100x
Experience or knowledge relevant to neural audio codecs, generative models, embedding systems, latent representations, multimodal speech-to-speech systems, model architectures, training schemes, inference algorithms, and hardware-efficient deployment

Benefits

Comp & perks

Remote work arrangement
AI-first work environment with encouragement to use and experiment with advanced AI tools
Opportunity to work on transformative voice AI research at scale
Open-source contribution and research publication opportunities
AI Notetaker interview recording can be declined without affecting candidacy