JOB-1892
A seed-stage startup that converts disorganized company data into structured insight is hiring an experienced applied-AI engineer to run the AI layer of its product. The company builds on hosted frontier models, and this person makes sure model behavior is measurable, dependable, trustworthy and helpful once live.
What you’ll own
- Retrieval and context assembly, structured extraction, and the choice of model and provider
- Orchestration of multi-step and tool-using workflows, plus monitoring of live systems
- Quality loops: representative datasets, evaluations, regression suites, safeguards and fallbacks
- Pipelines from raw customer inputs through to what users finally see
- Turning vague product goals into experiments, builds, metrics and releases
What you bring
- Four or more years of production software, ML or applied AI work, with reliable and observable AI services (latency, cost, retries and failures included)
- Hands-on LLM experience: structured outputs, retrieval, tool use and multi-step flows
- Python skills, and ideally Go or other backend experience
- Experience with evaluation methods such as curated reference sets, model-based grading, human review and launch gates
- Direct communication with customers and a startup or zero-to-one background
- Helpful extras: embeddings and reranking, document understanding, process or task mining, and daily use of AI coding tools
On-site in the San Francisco Bay Area, five days a week. Visa transfers considered.
