About the role

This role takes technical ownership of the core product: the end-to-end game generation pipeline. The person in this role will scale the system to drive up the playability pass rate while driving down the cost per generation at volume. Work spans the full AI generation loop: generation, validation, repair, and evaluation.

Responsibilities

- Own the end-to-end game generation pipeline from prompt to running browser game

- Build and maintain automated validation and repair loops for AI-generated code and assets

- Reduce cost per generation through tiered model routing, caching, distillation, and other LLM economics optimizations

- Integrate AI generation pipelines directly with web game runtimes (Three.js, Phaser, Babylon, WebGL)

- Work with procedural content generation systems for game mechanics and assets

- Lead a small team of 2-3 engineers on the AI generation stack

- Ship production code that compiles and runs complex AI-generated outputs in browser game environments

Requirements

- Shipped text-to-game, AI game-creation tech, or a prompt-to-running-app agent at scale

- Experience owning a full AI generation loop (generation, validation, repair, evaluation)

- Proven track record of cutting cost per generation in an LLM pipeline while maintaining output quality (via tiered model routing, caching, distillation, or similar)

- Hands-on engineering experience with web game runtimes (Three.js, Phaser, Babylon, WebGL) or procedural content generation

- Understanding of basic ML mathematics and LLM systems (not just prompt engineering or research)

- Experience beyond basic LLM wrappers; must have shipped production code that compiles and runs complex AI-generated outputs

- Not from a consultant, agency, or B2B employment setup; must be full-time