Prompt Engineer – Prompt Reliability Workstream

LumenWorks

Contract · Remote · Flexible hours

About the role

LumenWorks is revamping the performance and reliability of its flagship customer‑facing platform. As a Prompt Engineer you will lead the prompt design for the new "PulseCache" module, a high‑throughput caching layer that powers real‑time user interactions. Your work will directly impact latency, error rates, and overall service stability.

Key responsibilities

• Design, test, and iterate prompts that drive optimal model responses for latency‑critical API calls.
• Collaborate with SREs to embed prompt metrics into existing Grafana dashboards.
• Implement prompt versioning in GitLab CI pipelines and ensure rollback capability.
• Conduct A/B experiments on prompt variants to quantify performance gains.
• Document prompt schemas and best‑practice guidelines for the PulseCache team.
• Provide rapid troubleshooting support during the two‑week performance sprint.

Must-have skills

• Proficiency in crafting prompts for LLMs with a focus on speed and accuracy.
• Experience integrating prompts into Python and JavaScript microservices.
• Strong understanding of REST API design and latency optimization.
• Familiarity with monitoring tools such as Grafana, Prometheus, or Datadog.
• Ability to write clear technical documentation.

Nice to have

• Background in MLOps pipelines and model deployment.
• Exposure to generative AI safety and bias mitigation practices.