About the role
A small marketing services company in the United States is hiring for this role. You will own the design, versioning and documentation of prompts and agent workflows, and build the evaluation systems that show whether quality is holding up. You will work closely with engineering as the company's AI work moves into production.
What you'll do
- Design, version and document prompts and agent workflows
- Build evaluation systems that measure output quality
- Partner with engineering on retrieval pipelines, tool use and guardrails
- Run model benchmarking to compare options
- Monitor production performance and improve prompts over time
What you bring
- Hands-on experience designing and iterating on prompts and agent workflows
- Experience building evaluation systems for model output quality
- Working knowledge of retrieval pipelines, tool use and guardrails
- Comfort benchmarking models and tracking production performance
Details
- Location: United States
- Type: Full-time
Apply below and a Traverse recruiter will be in touch.