JOB-1911
A well-funded healthcare technology company is building AI products that listen to clinical conversations through phone and browser apps. It needs a staff-level backend engineer to make long-running, compute-intensive audio and inference pipelines dependable in an early, build-from-nothing setting.
The work
- Design, build and run the backend for the company’s audio products, including ingestion of lengthy recordings, media handling, and both live and asynchronous streaming
- Own orchestration of multi-step inference pipelines covering transcription, speaker separation, LLMs, clinical extraction and summarization, handling scheduling, queues, retries, timeouts, dead letters, idempotency and replay
- Collaborate with clinical, security, AI, product and mobile colleagues so workflows stay reliable on weak networks and offline
- Set service-level objectives and instrument workflows thoroughly with structured logs, metrics, tracing, dashboards and alerts
- Drive incident response and postmortems until recurring failures disappear
- Shape the APIs, data models, deployment pipelines and cloud infrastructure while balancing reliability, latency and cost, and mentor other engineers
Profile
- Seven to fifteen years in backend, platform or infrastructure roles (eight or more is typical)
- Production distributed systems built and operated at scale
- Python or a comparable modern backend language
- Currently or recently a staff engineer or tech lead, with a record of measurably improving reliability
- Exposure to speech, audio or live-streaming systems, or a healthcare background, is a plus
Hybrid in the South Bay or the San Francisco Bay Area, with at least three days on-site. Visa transfers are considered.
