Applied Scientist - Deep Learning

Role overview

Develop deep learning systems that improve the reliability of voice AI in live, demanding environments. The work covers audio-processing models, foundation-model adaptation, evaluation benchmarks, data pipelines, and experiments, combining applied research with strong engineering and a direct connection to customer outcomes. The position is fully remote within the United States and is not eligible for visa sponsorship.

• Train audio-processing deep learning models at different scales.
• Fine-tune and post-train foundation models for voice applications.
• Create benchmarks that reflect real customer and product requirements.
• Build data pipelines supporting model development and evaluation.
• Run experiments and use empirical results to improve system performance.
• Collaborate on applied science and engineering for production voice AI.

• PhD in a technical field or equivalent applied deep learning experience.
• Experience training deep learning models.
• Strong applied research and experimental problem-solving skills.

• Experience working with PyTorch.
• Background building neural audio-processing models.

• Fully remote role for candidates located in the United States.
• Distributed, collaboration-oriented work environment.
• Visa sponsorship is unavailable; candidates must be authorized to work in the United States.