About This Role
This fully remote position is open to candidates in 55 countries and focuses on evaluating and advancing AI capabilities in game creation and interaction. You will apply your expertise to help train next-generation AI systems, shaping how models learn, reason, and perform through high-quality, real-world input. No prior AI experience is required; your domain knowledge is what matters.
Responsibilities
- Design and build small-scale game projects or interactive scenes in a major game engine to serve as the foundation for assessment tasks
- Integrate realistic gameplay bugs that require thorough play testing and on-screen analysis to discover and resolve
- Develop multifaceted gameplay feature challenges that rely on multiple in-game systems and mechanics
- Draft clear and concise briefs for each task, emulating the style of instructions from players, designers, or producers
- Create comprehensive reference solutions—including both bug fixes and feature implementations—for every assignment
- Define detailed success criteria, including input sequences, expected visual outcomes, and verifiable in-game states
- Conduct end-to-end play testing of each task to guarantee reliability and solvability
Requirements
- Demonstrated experience in game development, whether in AAA, indie, solo, or serious hobby contexts
- A portfolio or verifiable published work (personal site, itch.io, Steam, GitHub, game jam entries, or shipped title credits)
- Advanced proficiency with at least one major game engine (Unity, Godot, Unreal), including editor tools, scene management, and debugging utilities
- Comfort with engine scripting languages such as C#, GDScript, C++, or Blueprints, with the ability to interpret, troubleshoot, and enhance gameplay logic
- Proven ability to identify and resolve gameplay issues through iterative play testing
- Strong written communication skills for drafting precise task instructions and acceptance criteria
Nice to have:
- Experience in game QA, technical design, systems design, automated testing, or prior contribution to AI evaluation projects
Important Details
- Work format: fully remote, open to candidates based in 55 countries, including United States, United Kingdom, Canada, Germany, India, France, Netherlands, Singapore
- Compensation structure: output-based; experts are paid per task that meets the project specifications
- Minimum submission requirements apply; experts must submit a minimum of tasks per week