Staff Applied AI Engineer, Product & Agent Performance

What Success Looks Like In 3 months:

• Establish a production-grounded baseline for priority agentic workflows with documented failure modes and evaluation rubrics.
• Map current retrieval, context, memory, and escalation patterns to identify highest-value improvements.
• Earn trust across Product and Engineering by turning production evidence into clear recommendations.

In 6 months:

• Deliver measurable improvements in accuracy, reliability, steerability, latency, or cost for one or more priority workflows.
• Validate human-review and escalation behavior under adversarial and edge-case conditions.
• Ensure production-representative evaluation suites and regression checks inform model-change decisions.

In 12 months:

• Build a repeatable product-layer AI performance practice that moves from production failure to diagnosis, experiment, evaluation, and release decision.
• Catch high-severity regressions earlier and make agent behavior more transparent, calibrated, and trustworthy at scale.
• Maintain model cards, intended-use guidance, limitations, and performance documentation that are current and useful to product and customer-facing teams.

What You'll Be Doing:

• Design and iterate on agent behavior across real, live workflows, including long-horizon, multi-turn agentic tasks.
• Design retrieval and context architecture so the right source data reaches a model in the right structure.
• Design memory and state handling across multi-turn and multi-agent flows, determining what is carried forward, summarized, or dropped and why.