RL Environments Specialist

About This Role

Create full Reinforcement Learning (RL) environments, including UI and backend, to programmatically generate and validate tasks for training computer-use agents, taking ownership of the entire task creation process for specific environments.

Responsibilities

• Build sandbox UIs that agents and RL actors will interact with
• Create tasks for built environments
• Programmatically validate task completion
• Take ownership of the entire task creation process for a given environment

Requirements

• Strong professional experience with React.js (hooks, modern state management, TypeScript preferred)
• Strong professional experience building backend services in Python (FastAPI, Flask, or Django)
• Hands-on experience with containerization (Docker required)
• Strong front-end design skills and exceptionally high taste in UI/UX, polish, and visual detail
• Proven ability to design a relational database schema in Python and populate it with large-scale, realistic mock data
• Experience creating and exposing clean, well-documented API endpoints (REST or GraphQL)
• Exceedingly high standards for code quality, readability, testing, and front-end craftsmanship
• Extensive day-to-day experience using coding agents / AI assistants as a power user (Cursor, Claude, Copilot, Grok, Aider, etc.)
• Good understanding of the Reinforcement Learning paradigm (RLHF, PPO, DPO, reward modeling, etc.)

Nice to Have

• Experience with Docker Compose/Kubernetes
• Posses strong logical reasoning skills, is detail-oriented, and thrives in a fast-paced work environment
• Eager to teach to and learn from teammates
• Enthusiasm to collaboratively build the best truth-seeking AI

Skills

* Required skills