PolyRook

AI 3D Environment Generation

Project Overview

PolyRook is my Startup Studio project at Cornell Tech that uses AI to generate fully customizable 3D environments for games and films—enabling creators to build and iterate immersive worlds in seconds. Our platform will provide 1-click generation of complete 3D scenes including terrain, props, and lighting with real-time collaboration capabilities.

Development Timeline

June 2024
Rapid Environment Generation (MVP) - Roll out to 2 pilot studios
Ship 5 ready-to-use biome presets (urban, forest, desert, tundra, sci-fi)
Instrument Key Metrics - Add hooks for Game Dev Time and AI Asset Usage
July 2024
Feedback & Quick Wins - 2-day feedback cycles with pilots
Expand Procedural Terrain - Rivers, coasts, & caves, "Biome mixer" UI
Real-Time Collaboration - Internal alpha → beta with tracking & locked edits
August 2024
Performance Optimization - LOD auto-generation and texture atlasing for props
Public v1.0 Environment Gen - Open self-serve sign up with launch video & docs
API & Plugin Kit - Unity & Unreal connectors with REST endpoints

Market Opportunity

Our total addressable market is $8.6B from the global 3D asset market (games, film, virtual production), with an initial serviceable obtainable market of $126M from the indie/SMB studios segment. We're targeting a growing need as 93% of surveyed studio leads expressed willingness to pay for 30% faster environment production.

Team

Zachary Decker
Co-founder
Robotics & Environment Generation
Jeremy Lu
Co-founder
Market Research & Business Development
Ryan Lewis
Co-founder
Game Development Background
Wentao Ye
Co-founder
Graphics Pipelines

Current Status

PolyRook is currently in Alpha stage with some active users but no revenue yet. We've been in active development for 6 months (since October 2024) and are preparing for our next phase of growth.