Summary
Raj Kolamuri is a mechatronics engineer specializing in robot perception and calibration for production-grade semiconductor tools, with eight years of experience delivering sub-millimeter, marker-free vision systems at the edge. At Lam Research he led the TREOS perception and calibration stack that achieved day-one auto-calibration and six months of zero field failures on tier-1 fab hardware, combining classical multi-view geometry with SIMD/Cython and CUDA performance work. His CMU research produced the first real-time in-hand rotation detection from vision-based tactile sensing, and taught him to prioritize system integration—calibration, synchronization, and failure modes—over isolated algorithms. Comfortable across Python, OpenCV, ROS, and learned vision, he focuses on problems where standard CV fails: tight optical regimes, non-cooperative targets, and hard latency budgets. He deliberately bridges classical CV and modern foundation-model approaches so he can own the full handoff from robust calibration and runtime to next-generation AI perception.
8 years of coding experience
2 years of employment as a software developer
Master of Science Research Mechanical Engineering, Master of Science Research Mechanical Engineering at Carnegie Mellon University
Bachelor of Technology - BTech(Hons.) Mechanical Engineering, Bachelor of Technology - BTech(Hons.) Mechanical Engineering at Indian Institute of Technology, Kharagpur
English, Telugu, Hindi