Work

Shipped systems, not slide decks

A selection of flagship builds that demonstrate what we bring to client engagements.

Flagship

GOAT RACER ONE

A fully autonomous RC racing platform trained entirely in simulation and deployed to real hardware. End-to-end Sim2Real: URDF authoring, IsaacLab environment, PPO training, ONNX export, TensorRT optimization, and Jetson deployment.

Demonstrated live on NVIDIA Omniverse Livestreams. The same loop we bring to client work: simulate, train, transfer, ship.

PPO
Proximal Policy Optimization, IsaacLab
ONNX
→ TensorRT inference pipeline
Jetson
Edge deployment, real hardware
Live
NVIDIA Omniverse stream feature
Capabilities

What the loop looks like

Every engagement moves through the same five-stage pipeline. We own all of it.

1 · Model the world

URDF / USD assets, sensor models, physics parameters. Digital twins from Gaussian splats or CAD.

2 · Train the policy

IsaacLab + PPO. Domain randomization, reward shaping, and curriculum across thousands of parallel envs.

3 · Export & optimize

PyTorch → ONNX → TensorRT. INT8 / FP16 calibration. Profile every stage of the graph.

4 · Deploy to hardware

Jetson Orin or DGX Spark edge. ROS 2 / custom runtime. Integrated with VESC, RealSense, and custom sensors.

5 · Measure & iterate

Telemetry, replay, and regression. Close the sim↔real gap with real-world data fed back to training.

Done · Hand off

Documentation, training for your team, and a reproducible pipeline so the loop keeps turning without us.

Got a system to build?

If you can describe the problem, we can scope the loop.

Book a call →