Senior Software Engineer (Agentic Coding RL Environments)

About the role

Surge AI partners with the world's leading AI labs to build the benchmarks and RL environments that train and evaluate frontier models. Most benchmarks cut corners because expert humans are expensive. Our model is the opposite: we pay experts enough that we don't have to.

Responsibilities

• Build coding RL environments and benchmarks that test how agents perform on realistic software engineering tasks
• Design tasks based on real engineering work
• Review and red-team model output to surface meaningful failures
• Write the tests, rubrics, and success criteria that separate good work from slop
• Give structured feedback that improves future models

Qualifications (Must-Have)

• Professional software engineering experience shipping real, scaled production systems
• Expertise in at least one common tech stack (TypeScript, Python, Ruby, PHP, Java, Rust, Go, or C++)
• Strong judgment about what "good code" and a "meaningful model failure" look like
• Comfort working independently, and giving direct, code-review-style feedback
• Availability for a meaningful commitment (most contributors work 20 to 40+ hours per week)

What Makes This Different

• No managers, 1:1s, OKR ceremonies, or Jira backlog grooming
• You choose when and where you work
• You are judged on the quality of what you produce, not speed

Application Process

• Coding capability screen (replaces a multi-round interview loop)
• ID verification
• Background check
• Paid first real task. Do well, and you're on the team.