JOBSEARCHER

Open Source Codebase Contribution

ARCHIVED

We can't find an active application page for this role right now. It may reopen or be listed elsewhere. Use Next Steps to search for an active apply link and similar live jobs.

https://discord.gg/5w2BNaQ6B7FastCrest is building Reflex, an open-source deployment and debugging layer for vision-language-action robot models. Reflex helps robotics builders take VLA policies like pi0, pi0.5, SmolVLA, and GR00T from research checkpoints to real hardware. The goal is to make the annoying parts easier: export, serve, validate, benchmark, trace failures, and run models on edge GPUs. This is an unpaid open-source contributor role, not a paid job or internship. It is for people who want to build in public, contribute to a serious robotics infrastructure repo, and work on real problems in physical AI. You will be moved to a paid role should you prove to be a valuable contributor. Good contribution areas: Add support for new VLA repos and model formats Improve ONNX / TensorRT export paths Build evals for robot policy behavior Add Jetson / edge GPU deployment tests Work on failure tracing, replay, and debugging tools Improve docs, examples, and contributor onboarding Explore ROS2, simulation, and real robot integration Useful background: PythonPyTorchRobotics, ML, or systems engineeringONNX / TensorRT / CUDA experience is a plus, but not required Comfort reading papers and turning them into working code If you want to contribute, apply with your GitHub, a short note on what part of the stack interests you, and any robotics, ML, systems, or open-source work you have done. Repo: https://github.com/FastCrest/reflex-vlaDiscord: https://discord.gg/5w2BNaQ6B7