ASR Engineer

About this role

Join an early-stage ambient intelligence startup as an ASR Engineer to own and evolve a cloud-based speech recognition and transcription pipeline. You'll drive measurable improvements in latency, accuracy, and reliability while collaborating across product, hardware, and international R&D teams in a lean engineering environment.

What you'll do

• Build and iterate the cloud-based ASR pipeline from audio capture through post-processing at production scale
• Own ASR quality and reliability end-to-end, shipping improvements on latency, small-word accuracy, and voice-print reliability
• Execute the full ML lifecycle: data preparation, model training and fine-tuning, evaluation, and deployment
• Translate lightweight product feedback into concrete, production-ready pipeline changes with minimal specification
• Collaborate with overseas R&D, hardware, and supply-chain teams across time zones
• Partner with product engineer on shared backend and pipeline infrastructure

What they're looking for

• ASR and transcription pipeline development
• Cloud-based audio processing and streaming pipelines
• ML lifecycle management (data handling, evaluation, deployment)
• Model training and fine-tuning
• Production ML debugging and optimization
• On-device or embedded ML frameworks (Core ML, TensorFlow Lite)
• Latency optimization and performance tuning
• Cross-functional collaboration in early-stage settings

Benefits

• Equity participation as an early engineering hire
• Hybrid work arrangement (3 days per week in San Francisco office)
• Direct partnership with product and general management leadership
• End-to-end ownership of critical infrastructure