Responsible for the architectural design and training of lightweight ASR models for mobile and embedded devices, minimizing parameter count and computational overhead while maintaining high recognition accuracy.
Lead the end-to-end optimization of ASR models using quantization, pruning, and knowledge distillation to meet stringent power, memory, and latency requirements on edge devices.
Manage the cleaning, alignment, and augmentation of multilingual speech data; address cold-start challenges for low-resource languages and enhance model robustness in complex, noisy environments.
Partner with engineering teams on model conversion.

Master’s or PhD in Computer Science, Signal Processing, or a related field, with 3–5 years of experience in speech algorithms. Must have a proven track record of deploying on-device or offline ASR models.
Expert-level command of PyTorch or TensorFlow, with deep proficiency in at least one mainstream E2E ASR framework (e.g., Wenet, Espnet, Icefall/K2, Zipformer) and a thorough understanding of their underlying architectures.
Demonstrated hands-on experience in model compression and Post-Training Quantization (PTQ) workflows, with the ability to independently resolve quantization-induced accuracy degradation.

Top-tier healthcare for employees and dependents, including dental and vision, and a generous employer subsidy.
401(k) plan for full time employees with company matching.
Unlimited PTO, plus 13 paid holidays.
12 weeks of paid time off to spend time with your new family, regardless of gender.
Minimum of 3x in office per week.
New hires are equipped with their choice of new top-of-the-line laptops and workstation setups.
Best office equipment. Annual offsites. Free office drinks and snacks.

Machine Learning Engineer – On-device ASR

Key skills