Core Simulation Platform Engineer (Open-Source)

United States Digital Space LLC; Paris (TX); On-site; USD 140,000 - 200,000; Full time

Job summary

the company-World is building a Python-first, open-source simulation platform for physical AI. Our engine runs on CUDA, ROCm, Metal, Vulkan and more, enabling fast, scalable robotics research and development.

As a core maintainer you will improve test infrastructure, reproducibility, and extensibility, collaborating with physics and compiler teams to create reliable, battle-tested tooling that empowers researchers and engineers.

Qualifications

• Strong Python engineering on large codebases with good design principles.
• Experience with performance profiling and debugging in complex systems.
• Open-source contribution and collaboration across teams.

Responsibilities

• Build scalable testing for a JIT-compiled engine to reduce runtime bottlenecks.
• Make bugs reproducible across machines and backends with automated repro scripts.
• Advance autodiff integration and explainable performance improvements.
• Improve scene editing and insulation from full builds for faster iteration.
• Develop extensible plugin architecture for solvers and viewers.
• Ensure maintainability, typing discipline, and robust CI pipelines.
• Collaborate with physics and compiler teams to drive integration.

Skills

Python

Code architecture

Performance profiling

Open source

Tools

PyTorch

CUDA

Metal

RenderDoc

Job description

the company-World is building a Python-first, open-source simulation platform for physical AI. Our engine runs on CUDA, ROCm, Metal, Vulkan and more, enabling fast, scalable robotics research and development.

As a core maintainer you will improve test infrastructure, reproducibility, and extensibility, collaborating with physics and compiler teams to create reliable, battle-tested tooling that empowers researchers and engineers.