Summary
Ujas Mandavia is a Senior Software Engineer specializing in perception and autonomy with eight years of experience building SLAM, VIO, and sensor-fusion systems for robotics and autonomous platforms. He combines deep expertise in multiview geometry, 3D point cloud processing, non-linear optimization, and probabilistic filtering with practical experience integrating IMUs, cameras, and LiDAR in real-time systems. Across roles at OKSI, Forterra, Magna, and Thales, he has moved research-grade algorithms into production, shipping monocular and stereo VO, LiDAR–camera fusion, and mapping/localization stacks. Comfortable in C++ and Python, he leverages PyTorch/TensorFlow and ROS to bridge classical geometry with learning-based components for geometric tasks. Based in Boston, he pairs a Northeastern MS in robotics with a maker’s mindset and a penchant for exponential thinking, often blending hands-on hardware work with algorithmic innovation. Notably, his background includes deploying on constrained platforms (Turtlebot-3/RaspiCam) and real-time fusion of heterogeneous sensors for robust fielded SLAM.
8 years of coding experience
6 years of employment as a software developer
Bachelor of Engineering - BE Mechanical Engineering Mechatronics, Bachelor of Engineering - BE Mechanical Engineering Mechatronics at Gujarat Technological University (GTU)
Master of Science - MS Robotics Computer Vision Machine Learning Industrial Engineering, Master of Science - MS Robotics Computer Vision Machine Learning Industrial Engineering at Northeastern University