MeshEvolver - Procedural 3D Generation

Technical Overview

MeshEvolver is an experimental system for procedural generation and 3D visualization. It combines a Python backend with FastAPI, Redis, and Pydantic, a Rust core exposed through PyO3 for geometry computation, and a React 19 frontend with Three.js and React Three Fiber. The project demonstrates how to separate heavy computational logic, web orchestration, and visual exploration in a multi-language architecture.

Problem Statement

Procedural generation requires iterating parameters, calculating variants, and visualizing results quickly. A single runtime is not always ideal for combining performance, web APIs, and an interactive 3D experience.

Architecture

Layered architecture: Vite frontend with React 19, Three.js, React Three Fiber, and Tailwind CSS 4; FastAPI backend with Redis, Pydantic, and NumPy; and a Rust `cdylib` crate with PyO3, `noise`, `rand`, `serde`, and `serde_json`. Rust concentrates geometry computation, Python orchestrates services, and React renders the 3D experience.

Key Features

• 3D frontend with Three.js and React Three Fiber
• FastAPI backend for orchestration and endpoints
• Rust core integrated with Python through PyO3
• Dependencies for procedural noise and serialization
• Redis as a cache or data coordination layer
• UI controls for exploring parameters and visual results

Challenges

• Separating responsibilities between Rust, Python, and TypeScript
• Designing a maintainable PyO3 bridge for procedural computation
• Rendering interactive 3D geometry in the browser
• Maintaining data consistency across API, core, and client

Outcomes

Solid technical prototype for experimenting with procedural generation, integrating Rust computation, Python APIs, and WebGL visualization in one platform.