About the RoleYou'll own the transcription pipeline end-to-end at an early-stage ambient intelligence startup - one of the company's first US engineering hires. Working directly with the GM and Head of Product, you'll build, tune, and ship improvements to a cloud-based ASR system with a narrowly scoped on-device component, making real tradeoffs between latency, accuracy, and reliability as the product evolves.
What You'll Do- Build and iterate on the cloud-based ASR pipeline, from audio capture through post-processing, in production at scale.
- Own ASR quality and reliability end-to-end, shipping measurable improvements in latency, small-word accuracy, and voice-print reliability.
- Work across data, training/fine-tuning, evaluation, and deployment to translate product feedback into shipped pipeline changes.
- Collaborate with hardware and R&D teams across time zones.
- Partner with a Product Engineer on shared backend and pipeline surfaces.
- Operate with minimal specification, turning lightweight asks into concrete, production-ready improvements.
What We're Looking For- 3+ years building and tuning transcription/ASR pipelines end-to-end in production, primarily in cloud-based settings.
- Demonstrated ownership of production ASR systems across the full lifecycle: data preparation, model training/fine-tuning, evaluation, and deployment.
- Experience building and optimizing latency-sensitive or streaming audio/ASR pipelines.
- Track record of shipping pipeline improvements from design through deployment and ongoing iteration based on real usage data.
- Experience debugging transcription quality issues in production (small-word accuracy, voice-print reliability, latency).
- Comfort in early-stage or founding engineering environments with minimal specs and small teams.
- On-device or embedded ML experience (Core ML, TensorFlow Lite, or similar frameworks) is a strong plus.
- Prior experience building wearable, hardware, or robotics device products is a plus.
- Background at AI-native consumer applications focused on transcription or audio is a plus.
- Experience building agent or LLM-based product features (tool use, memory, retrieval) is a plus.
- You care how transcription feels to use, not just how it benchmarks.
LocationHybrid (3 days/week in office) - San Francisco Bay Area, California. Visa sponsorship is not available.
