RL Environments Architect

As an RL Environments Architect at Surge, you will design and govern simulated worlds where AI agents learn, defining primitives, reward structures, and telemetry to ensure training signals are stable and scalable. You will establish standards for data quality and reproducibility across large-scale agent ecosystems while partnering with researchers to translate real-world tasks into robust simulations. This role is critical for building the backbone of safe, rapid progress in agentic systems by creating modular frameworks and mitigating issues like reward hacking.