About the Role
We are looking for a Senior or Staff Software Engineer with a strong background in Visual Odometry (VO) and Localization to join our autonomy team. This role is critical to building precise, real-time vehicle localization systems using camera-based perception for autonomous driving applications. The ideal candidate has hands-on experience deploying VO systems in real-world environments, particularly in the autonomous vehicle (AV) domain, and brings solid C++ engineering skills to architect robust, high-performance software solutions.
This role is onsite 5 days per week at our Mountain View, CA office.
What you'll do
Design and develop real-time Visual Odometry pipelines using monocular, stereo, or RGB-D camera inputs.
Implement robust camera-based localization algorithms, including visual-inertial odometry (VIO), feature tracking, motion estimation, and scale recovery.
Integrate VO systems with IMU, GPS, and other sensor data to enhance pose estimation accuracy and stability.
Collaborate with the mapping, perception, and control teams to integrate localization with the AV software stack.
Develop and optimize production-quality code in modern C++ for real-time performance on embedded compute platforms.
Analyze system performance in diverse environmental conditions and drive improvements for reliability, accuracy, and robustness.
Participate in code reviews, mentor team members, and contribute to architectural decisions.
Stay up to date with the latest research in SLAM, VIO, and VO, and help transition promising techniques into production
What we're looking for
Bachelor’s or Master’s degree in Computer Science, Robotics, Electrical Engineering, or a related field.
5+ years of experience in robotics, computer vision, or autonomy; 3+ years specifically in Visual Odometry or VIO.
Experience working on Autonomous Vehicle platforms (e.g., development, testing, or deployment of AV systems).
Expert proficiency in C++ (C++14/17/20) and modern software engineering best practices.
Solid understanding of epipolar geometry, camera calibration, bundle adjustment, and optimization techniques.
Hands-on experience with open-source VO/SLAM libraries such as ORB-SLAM, VINS-Fusion, OpenVINS, or similar.
Experience working with ROS/ROS2, Linux development environments, and version control systems.
Bonus Qualifications
Experience integrating visual odometry with Lidar, GPS, or map-based localization.
Knowledge of GPU acceleration techniques (CUDA/OpenCV/OpenGL) for computer vision pipelines.
Familiarity with real-world deployment constraints such as environmental variability, sensor degradation, and compute limitations.
Experience in sensor calibration, time synchronization, and data preprocessing pipelines.
Contributions to relevant open-source projects or publications in VO, VIO, or SLAM.