Research Staff, Voice AI Foundations

📋 Description

• • As a Research Staff member in Voice AI Foundations at Deepgram, you will pioneer Latent Space Models (LSMs) to solve fundamental data, scale, and cost challenges in voice AI, enabling universal human-machine interaction through advanced audio understanding and generation.
• • Your day-to-day work will involve researching and implementing next-generation neural audio codecs, steerable generative models for diverse speech synthesis, embedding systems for interpretable latent space factorization, latent recombination for synthetic data generation at scale, and hardware-adapted model architectures for efficient training and real-time inference on massive audio datasets.
• • You will join a world-class research team at Deepgram, the leading platform powering the trillion-dollar Voice AI economy, where over 200,000 developers and 1,300+ organizations build voice applications using Deepgram’s APIs, including industry leaders like Twilio, Cloudflare, and NVIDIA-backed innovations.
• • In this role, you will push the boundaries of audio AI by developing transformative techniques that enable training on billion-hour datasets, real-time inference for hundreds of millions of concurrent conversations, and empathic, human-like speech-to-speech systems — advancing the state of the art while gaining deep expertise in foundation models, scalable systems, and AI-first research practices.