Senior Software Engineer, RL Environments

5 months ago

Responsibilities

Design, build, and refine reinforcement-learning tasks from ideation through grading, failure analysis, and iteration.
Own complex environments involving multi-step workflows, realistic stakeholder interactions, large codebases, technical debt, and system-design problems.
Direct coding agents in day-to-day work, critically evaluate their output, and identify subtle failures.
Distinguish model capability gaps from grader or environment issues and redesign tasks to target deeper failure modes.
Contribute to shared infrastructure and tooling used by the environments team.
Mentor newer engineers as the team grows.

Deep software-engineering experience across multiple domains, with expertise in infrastructure, distributed systems, performance, security, compilers, databases, or a similar specialty.
Proficiency in Python.
Extensive hands-on experience with coding agents, including Claude Code, Cursor, Codex, or similar tools.
Strong intuition for model behavior and the ability to design tasks around model shortcuts and failure modes.
Comfort working independently on complex, ambiguous problems with minimal direction.
A track record of owning work end-to-end.
Experience as a senior or staff engineer at an engineering-rigorous company, deep specialty expertise in an area current models struggle with, early-stage startup engineering experience, or substantial work with coding agents and agent evaluation may be a good fit.

Competitive cash and equity compensation, with compensation above the 90th percentile.
Health, vision, and dental benefits.
401(k) match.
Visa sponsorship and relocation support available.
Ownership and autonomy in a fast-moving startup environment.
Opportunity to work with leading machine learning engineers.

11-50 employees

Preference Model builds RL environments that automate ML research and engineering.