Autonomy Droid Perception SWE - Offboard Systems
Core
Design, train, and deploy cloud-side computer vision ML models to support and scale on-device perception for Zipline's autonomous drone delivery system.
Role type
Senior/Staff Perception Engineer (Offboard Systems)
Builds
Cloud-based perception models, semantic segmentation, 4D geometry, and preference modeling systems for autonomous logistics.
Domain
Autonomous logistics / Drone delivery / Computer Vision / Remote Sensing
Deliverable
production ML models
Required skills
Deep learning perception systems, 3D geometry, semantic understanding, mapping from remote sensing data, classical computer vision, CNN/Transformer architecture optimization, training and evaluation pipelines, distributed training debugging
Preferred skills
Experience deploying models in real-world robotics or autonomy applications, building staged roadmaps for challenging ML tasks
Technologies
CNN, Transformer, Gaussian Splatting, GPU clusters
Responsibilities
Design and implement computer vision ML models running in the cloud; Train and deploy large-scale models for semantic segmentation and 4D geometry; Design tools to predict deliverability and generate high-fidelity priors; Design evaluation and validation infrastructure for field reliability; Ship new features or models to production every 6 weeks; Lead architectural decisions and experimentation (Senior) or own roadmapping for offboard models (Staff)
Seniority
Senior (5+ years) / Staff (8+ years)
