Computer Vision Software Engineer, Software Defined Weapon Optic

ABOUT THE TEAM

Our Anduril Warfighter Systems division is building the next generation of precision engagement systems. We are developing a software-defined weapon optic that fuses real-time computer vision, sensor data, and onboard processing to dramatically increase first-round accuracy for the warfighter. By pushing advanced CV algorithms into ultra-constrained compute environments, we are delivering a decisive edge at the tactical level — including enabling counter-UAS engagements and other precision targeting missions. This team operates at the intersection of robotics, embedded systems, and defense — building technology that saves lives.

ABOUT THE JOB

Come join a high-performing team dedicated to fielding next-generation precision targeting systems. You will solve challenging computer vision problems that arise in a software-defined weapon optic providing precision aiming and targeting — including target detection, tracking, camera/sensor calibration, and geometric pose estimation — all running on resource-constrained embedded platforms. This role requires deep technical expertise in real-time vision algorithms, creative problem-solving under tight compute and power budgets, and a passion for delivering capability to the warfighter.

WHAT YOU WILL DO

The Computer Vision Software Engineer will focus on developing and optimizing vision algorithms for a software-defined weapon optic that augments the warfighter with precision aiming and targeting, including counter-UAS applications. Core responsibilities include:

• Designing and developing real-time target detection, classification, and tracking models and algorithms optimized for low-compute, embedded environments (e.g., Qualcomm-based platforms).
• Building and refining camera and sensor calibration pipelines (offline and online) for multi-sensor weapon sight assemblies, including thermal and electro-optical sensors.
• Developing pose estimation and localization algorithms using techniques such as bundle adjustment, feature detection/matching, homography fitting, and multi-view geometry.
• Profiling, optimizing, and deploying vision models and algorithms on constrained hardware — managing trade-offs between accuracy, latency, and power consumption.
• Creating testing and evaluation frameworks: designing experiments, analyzing field and production data, and establishing metrics for detection/tracking and overall algorithm performance under operational conditions.
• Collaborating closely with firmware, mechanical engineering, and manufacturing teams to integrate vision software into production weapon sight hardware.

REQUIRED QUALIFICATIONS

• BS in Robotics, Computer Science, Electrical Engineering, Mechanical Engineering, Mechatronics, or related field.
• Proficiency in C++ and Python.
• Strong understanding of fundamental computer vision concepts: camera projection models, image distortion, rigid transformations, multi-view geometry.
• Prior experience with geometric computer vision (SLAM, camera calibration, stereo vision, photogrammetry, or similar) in academic or professional work.
• Experience developing or deploying algorithms in resource-constrained or embedded environments.
• Good communication skills for effective collaboration across hardware, software, firmware, and manufacturing teams.
• Eligible to obtain security clearance.

PREFERRED QUALIFICATIONS

• MS or PhD in Robotics, Computer Science, Electrical Engineering, Mechanical Engineering, Mechatronics, or related field.
• Experience with industrial computer vision systems, counter-UAS applications, or precision targeting systems.
• Hands-on experience with Qualcomm processors/platforms (e.g., Snapdragon) or similar embedded vision SoCs.
• Experience with multi-camera and multi-sensor (EO/IR) systems and associated calibration techniques.
• Proficiency with mathematical/optimization packages in C++ and Python (e.g., Ceres, Eigen, OpenCV, NumPy, SciPy).
• Experience training, optimizing, and deploying detection/tracking models with TensorRT, SNPE, ONNX Runtime, or similar inference frameworks for edge deployment.
• Familiarity with real-time operating systems (RTOS) or embedded Linux environments.
• Familiarity with numeric computing platforms such as MATLAB, Octave, or Jupyter.