Perplexity
Member of Technical Staff (Software Engineer, Enterprise Agent Evals & Efficiency)
$220K–$405K
San Francisco +2 · Staff · Full-time · Posted Oct 19, 2026
- Location
- San Francisco
Seattle
New York City - Experience
- 5+ years of software engineering or machine learning engineering experience.
About the job
Build evaluation environments and infrastructure for enterprise AI agents, diagnose failures, and ship improvements that increase task success while reducing latency and cost.
What you'll do
- Create realistic, reproducible enterprise scenarios and evaluation suites with automated graders.
- Analyze agent performance and production failures, then improve agent harnesses, tools, skills, and model routing.
- Improve agent efficiency and reliability, and build regression gates that verify gains in production.
What they ask for
- 5+ years of software engineering or machine learning engineering experience owning production systems.
- Strong system design skills and proficiency in a backend language such as Python, Go, or Rust.
- Hands-on experience building, evaluating, or improving LLM-based applications or agents, including tool use and multi-step workflows.
- Deep expertise in at least one listed area: evaluation infrastructure, agent harnesses and tools, multi-system test environments, or model selection and inference efficiency.
Pay
$220K – $405K (base salary)
Tools
Python, Go, Rust
AI focus
Build and evaluate enterprise agents, including tool use, multi-step workflows, evaluation and grading, model selection and routing, context management, and inference efficiency.
Model-written summary · the original posting is authoritative