3 hours ago
San Francisco, CA, USAMid Level
Base Salary
$150k - $200k/yr
Responsibilities
- Build and iterate on a cloud-based ASR pipeline in production at scale from audio capture through post-processing.
- Own ASR quality and reliability while improving latency, small-word accuracy, and voice-print reliability.
- Handle data preparation, model training and fine-tuning, evaluation, and production deployment.
- Translate product feedback into shipped pipeline improvements and make latency, accuracy, and reliability tradeoffs.
- Collaborate with product, R&D, hardware, supply-chain, and backend engineering partners across time zones.
- Turn loosely defined requirements into concrete, production-ready improvements.
Requirements
- At least 3 years of experience building and tuning transcription and ASR pipelines end-to-end in production, primarily in cloud-based environments.
- Experience with latency-sensitive or streaming audio and ASR pipelines.
- Proficiency across data handling, model training and fine-tuning, evaluation metrics, and production deployment.
- Experience making production tradeoffs among latency, accuracy, and reliability based on user feedback.
- Comfort working in early-stage or founding engineering environments with minimal specifications and small teams.
- On-device or embedded ML experience with Core ML, TensorFlow Lite, or similar frameworks is a strong plus.
- Experience with wearable, hardware, or robotics device products is a plus.
- Background in AI-native consumer applications focused on transcription or audio is a plus.
- Experience building agent or LLM-based product features involving tool use, memory, or retrieval systems is a plus.
Benefits
- Base salary of $150,000 to $200,000 USD annually plus equity participation.
- Hybrid work arrangement requiring 3 days per week in the San Francisco Bay Area office.
- Visa sponsorship is not available.