Research Engineer Audio and Speech

Skills

About the Role

You will design streaming speech agent harnesses and train multimodal speech models. You will improve speech recognition and generation, evaluate production calls, and optimize inference for accuracy, latency, responsiveness, throughput, stability, and cost.

Requirements

• 2+ years of experience in speech, audio ML, multimodal ML, or production machine learning
• Experience developing or adapting autoregressive, diffusion, flow-matching, or codec-based speech models
• Experience with streaming agent systems, low-latency inference, production model serving, and real-world audio evaluation
• Python fluency
• Experience with a modern deep-learning framework such as PyTorch
• Knowledge of machine learning and signal processing
• Track record of taking research ideas from prototype to measurable production impact

Responsibilities

• Design and build streaming speech agent harnesses
• Research and train multimodal and full-duplex speech models
• Improve speech recognition, voice activity detection, endpointing, and speech generation
• Build evaluations and use production calls to improve accuracy, latency, naturalness, and task outcomes
• Optimize end-to-end inference and partner to deploy systems at scale

Benefits

• Medical, dental, and vision benefits for employees and their families
• Life insurance and disability benefits
• Retirement plan
• Parental leave
• Fertility and family-building benefits through Carrot
• Monthly wellness, lifestyle, and work-life balance stipend
• Daily office lunches and snacks
• Flexible vacation policy