Role overview
We're looking for someone to own the perception side of our stack — the part that figures out where real objects are so our XR developers can anchor virtual content to them.
You'll work alongside our Unity XR developers. They render the AR experience; you make sure the world model they're rendering against is accurate, stable, and updated in real time. We're primarily working with LiDAR today, and the object tracking pipeline is where our product stands or falls.
This is a hands-on engineering role, not a research role. What we need is someone who can take a messy real-world sensor stream and turn it into a clean, stable transform that Unity can trust.
What you'll work on
- Build object detection and 6DoF pose tracking for industrial equipment — both stationary machines and moving parts
- Work with LiDAR point clouds: registration, segmentation, fusion with RGB where it helps
- Keep AR anchoring stable — debug jitter, drift, and coordinate-frame issues against real hardware
- Set up the data collection and labeling pipeline so we can train models for new object classes as we take on new clients
- Train, evaluate, and iterate on detection/pose models (PyTorch or TensorFlow)
- Hand off clean interfaces to the Unity team — you don't need to write Unity, but you need to understand what the Unity side consumes
- Optimize for real-time performance on the hardware our clients actually run
What we're looking for
- LiDAR or depth sensor experience (Velodyne, Ouster, Intel RealSense, iPhone/iPad LiDAR, etc.)
- ICP and point cloud registration
- ONNX deployment, CUDA, or edge inference optimization
- Unity integration experience (on either side of the handoff)
- Industrial or production-environment work (factory, warehouse, plant)
- Experience building data collection and annotation pipelines
While this is initially unpaid, exceptional performance may lead to a paid position after the internship.
If you’re excited about this opportunity and eager to learn, please send your resume outlining your interest in the position. We look forward to helping you grow your skills and advance your career in backend development.
Job Types: Full-time, Internship
Application Question(s):
- I confirm that I am aware the role is unpaid.
- How many years of experience with PyTorch or TensorFlow do you have?
Experience:
- Python programming: 2 years (Preferred)
- C++ development?: 1 year (Preferred)
- computer vision (e.g., OpenCV): 2 years (Preferred)
Work Location: Remote