Shield AI
Manager, Software - Perception (R3770)
Washington, D.C. · Manager · Full-time · Posted Sep 28, 2026
- Experience
- Typically 7 years with a Bachelor’s degree; 5 years with a Master’s degree; 4 years with a PhD; or equivalent work experience. Also requires 5+ years in Unmanned Systems programs in the DoD or applied R&D and 2+ years of people leadership.
- Education
- BS/MS in Computer Science, Electrical Engineering, Mechanical Engineering, Aerospace Engineering, and/or similar degree, or equivalent practical experience.
About the job
Leads a software team developing and integrating real-time perception systems for autonomous aircraft, including object detection, sensor fusion, and state estimation.
What you'll do
- Lead multidisciplinary teams and coordinate technical execution across autonomy, integration, and testing.
- Develop and optimize perception algorithms for object detection, classification, tracking, sensor fusion, and state estimation.
- Integrate and validate perception software with sensors, autonomy systems, and onboard compute in simulation and operational settings.
What they ask for
- Degree in a listed or similar field, or equivalent practical experience.
- Typically 7 years of related experience with a Bachelor’s degree, 5 years with a Master’s degree, or 4 years with a PhD; equivalent work experience is also stated.
- 5+ years of experience in Unmanned Systems programs in the DoD or applied R&D, and 2+ years of people leadership experience.
- Ability to obtain a SECRET clearance.
Nice to have
- Hands-on integration or algorithm development with airborne sensing systems.
- Experience with PyTorch or Tensorflow, deploying on SWaP-constrained platforms, or validating systems during flight tests.
Tools
Kalman Filters, PyTorch, Tensorflow, SLAM, IMU, GPS, EO/IR
AI focus
Develops perception algorithms for autonomous aircraft, including deep learning-based object detection, classification, multi-target tracking, sensor fusion, and state estimation.
Model-written summary · the original posting is authoritative