Staff Deep Learning Engineer, State Estimation (R5785)
Core
Build deep learning capabilities for vision-based navigation to help autonomous systems localize and understand motion when GPS is unavailable.
Role type
Staff Deep Learning Engineer (State Estimation & Vision)
Builds
Learned components for vision-based navigation integrated into the Hivemind SDK for autonomous aircraft.
Domain
Defense technology, autonomous systems, 3D computer vision, geometric estimation.
Deliverable
production ML models
Required skills
Deep learning, 3D computer vision, projective geometry, multi-view geometry, Python, PyTorch, data pipeline engineering, model profiling, experiment tracking
Preferred skills
Aerial imagery, geospatial data, model export/quantization (TensorRT/ONNX), embedded deployment, physical platform validation, diffusion models, flow matching
Technologies
PyTorch, Python, TensorRT, ONNX
Responsibilities
Develop and evaluate models for feature detection, matching, depth estimation, and localization; combine learned representations with geometric methods; own data preparation and supervision strategies; build reproducible training workflows; design evaluations for model performance under varying conditions; analyze failures and prioritize improvements; partner on integrating learned measurements into navigation systems; profile models for onboard compute constraints; deliver documented components for the SDK.
Seniority
Staff, hands-on IC with strategic ownership
