Senior RL (Reinforcement Learning) Engineer - San Francisco, CA

Senior Reinforcement Learning EngineerSan Francisco, CA (FiDi) | On-site 5 days | Full-time$300,000 -- $500,000 + Equity A small, elite AI research team working on reinforcement learning in open-ended settings. The team includes researchers from leading PhD programmes and tier-one AI organisations. Early-stage, well-resourced, and moving quickly -- this is a genuine ground-floor opportunity with significant scope and impact.The RoleWe are hiring for a Senior RL Engineer to sit at the intersection of research and production engineering. You will own the translation of RL research ideas into reliable, measurable training systems and drive technically complex projects end to end.Key Responsibilities

• Build and improve RL training pipelines for language model-based agents
• Implement reward functions, verifiers, environment interfaces, rollout pipelines, and evaluation harnesses
• Design experiments to test whether RL methods are improving model behaviour, sample efficiency, robustness, or generalisation
• Build monitoring tooling: regression tests, eval suites, and reward-hacking checks
• Debug unstable training runs and diagnose learning dynamics failures across algorithms, rewards, data, infrastructure, and evals
• Manage GPU clusters, distributed training, and compute efficiency
• Build 0-to-1 systems for new RL workflows and harden them into reusable infrastructure
• Own ambiguous technical problems from problem framing through to delivery

Requirements

• Strong applied ML engineering background: shipped systems, open-source work, competitions, or early-stage startup experience
• Hands-on experience scaling RL pipelines and debugging training issues
• Familiarity with RL environments and large language models; diffusion model experience a plus
• Python proficiency and strong working knowledge of PyTorch or JAX
• Solid grounding in RL, supervised learning, optimisation, and modern deep learning
• Independent, intellectually curious, and able to drive ambiguous problems to working solutions
• Comfortable collaborating with researchers while holding high engineering standards
• PhD not required -- strong applied experience equally valued

Package

• Visa sponsorship available (H1B transfer, TN, OPT, O-1); existing US work authorisation preferred