Shield AI
Staff Software Engineer, Perception (R5421)
$195K–$293K
Washington, D.C. +4 · On-site · Staff · Full-time · Posted Jul 28, 2026
- Location
- Washington, D.C.
Dallas, Texas
San Diego, California
Boston, Massachusetts
Seattle, Washington - Experience
- Typically a minimum of 7 years of related experience with a Bachelor’s degree; or 6 years and a Master’s degree; or 4 years with a PhD; or equivalent work experience.
- Education
- Bachelor’s degree; or Master’s degree; or PhD; or equivalent work experience.
About the job
Leads development, evaluation, and deployment of vision, vision-language, and vision-language-action models for autonomous systems, translating machine-learning research into production-ready perception capabilities.
What you'll do
- Design, train, fine-tune, and maintain models and scalable training, data, and evaluation workflows.
- Deploy and optimize models for embedded hardware and integrate them into autonomous systems.
- Lead technical architecture and model development, collaborate across teams, and mentor engineers.
What they ask for
- Expertise in machine-learning fundamentals and production computer-vision model training and deployment.
- Strong understanding of 3D vision problems and algorithms.
- Experience with PyTorch and TensorFlow; demonstrated model deployment expertise using TensorRT and ONNX.
- Proficiency in C++ and Python; ability to obtain a SECRET clearance.
Nice to have
- Experience developing autonomous systems for defense customers or training/fine-tuning vision-language, vision-language-action, or world models.
- Open-source contributions or publications in leading computer-vision and robotics venues.
Pay
USD 195,200–292,800 per-year-salary (base salary)
Tools
PyTorch, TensorFlow, TensorRT, ONNX, C++, Python
AI focus
Develops vision, vision-language, and vision-language-action models for autonomous-system perception and decision-making, including data pipelines, supervised fine-tuning, evaluation, and deployment.
Model-written summary · the original posting is authoritative