Game Development Expert (LLM Post-Training Data & Environments)

What you’ll do

1. 01Training task design: develop tasks based on real-world workflows for AI game development, including gameplay implementation, scene building, asset integration, interaction debugging, and performance optimization. Define each project’s starting state, requirements and constraints, deliverables, and acceptance criteria, and build task sets spanning different difficulty levels.
2. 02Expert data production: independently complete representative tasks, capturing runnable projects, code and asset changes, tool operation logs, and debugging processes. Diagnose issues in model outputs, demonstrate corrections, and provide preference feedback to create expert data for model post-training.
3. 03Development environments and tools: build game-engine and content-creation environments that models can operate. Develop scripts, plugins, and tool interfaces for loading projects, editing scenes, building and running projects, capturing state, and resetting environments, ensuring tasks are executable and processes reproducible.
4. 04Quality standards and verification: establish acceptance criteria covering requirements compliance, functional correctness, playability, visual quality, and runtime performance. Verify deliverable quality through automated tests, runtime results, and expert review, and develop verification mechanisms for training feedback and evaluation.
5. 05Model failure analysis and data iteration: analyze model failures in game development and distinguish between gaps in development capabilities, tool-use problems, and environment issues. Work with algorithm and data teams to add targeted tasks and feedback data, and track improvements in model performance.

What we look for

1. 01A bachelor’s degree or higher in computer science, software engineering, game development, digital media, or a related field, with 3–5 years of experience in game development or game toolchain R&D and experience contributing to the development and delivery of a complete game or complex interactive project.
2. 02Proficiency in at least one game engine, such as Unreal Engine or Unity, with the ability to independently complete the development process from requirements breakdown and implementation through debugging and acceptance testing; familiarity with content-creation tools such as Blender and workflows for importing and integrating assets, including models, materials, and animation.
3. 03Proficiency in the development languages used by your engine, such as C++ or C#, and the ability to develop tools and automate workflows with Python or another scripting language, along with solid version-control and troubleshooting skills.
4. 04The ability to turn open-ended game development requirements into well-scoped tasks of appropriate difficulty with verifiable outcomes, accurately assess the quality of code, scenes, and runtime behavior, and recommend concrete, actionable fixes.
5. 05A sustained interest in LLMs and AI-assisted game development and a willingness to turn development experience into tasks, data, and quality standards, along with rigorous process documentation habits and cross-team collaboration skills.

Nice to have

• Experience with engine editor extensions, Blender plugins, procedural content generation, asset pipelines, or automated builds and testing.
• Hands-on experience developing games with Coding agent tools, or contributing to LLM training data, agent tool environments, or evaluation benchmarks.
• Playable games, engine plugins, open-source projects, or technical work that demonstrates the full development process and key technical decisions.