Fuku
Senior Prompt Engineering Engineer
San Francisco, California · Senior · Full-time · Posted Mar 2, 2026
- Experience
- 4+ years shipping ML/AI or applied NLP systems, including at least 1–2 years focused on LLM prompt or agent engineering.
About the job
Owns the design, evaluation, tooling, and production rollout of prompt- and agent-driven workflows for enterprise AI products.
What you'll do
- Design, test, and ship multi-agent and multi-turn prompt strategies.
- Build automated evaluation, red-teaming, and telemetry systems to measure quality, safety, and latency.
- Integrate tools and knowledge grounding, productionize reusable components, and collaborate with product teams and customers.
What they ask for
- Familiarity with Claude, GPT, Gemini, or open-source model families and evaluation tools such as Ragas, DeepEval, LangSmith, or Weights & Biases.
- Strong software fundamentals in Python or TypeScript, plus workflow orchestration experience with LangChain/LangGraph, LlamaIndex, custom evaluation harnesses, or similar tools.
- Comfort collaborating with customers, product, and go-to-market partners; strong written documentation habits.
- A bias toward experimentation and measurable outcomes.
Nice to have
- Experience deploying AI copilots in regulated industries, multilingual surfaces, or high-volume support funnels.
- Background in retrieval, structured reasoning, toolformer-style call planning, synthetic data pipelines, or open-source LLM tooling.
Tools
Claude, GPT, Gemini, Ragas, DeepEval, LangSmith, Weights & Biases, Python
AI focus
Designs and productionizes enterprise LLM prompt and agent workflows, including automated evaluation, red-teaming, guardrails, tool integration, and model benchmarking.
Model-written summary · the original posting is authoritative