Mohamed R.

I’m a Computer Vision and Machine Learning Engineer with over 3 years of experience in PyTorch, CUDA programming, and 3D reconstruction. My focus is on building high-performance vision and AI systems — from Neural Radiance Fields (NeRF) and Gaussian Splatting to sensor fusion, and SLAM. I hold an M.Sc. in Geodesy and Geoinformation Science (Computer Vision specialization) from TU Berlin, and my work bridges theory and implementation. I’m skilled at turning complex research papers into optimized, production-ready code that runs efficiently on GPU architectures. I regularly work with CUDA kernels, Tensor Cores, and GPU-based parallelization to accelerate model training and inference. Key Skills: CUDA programming and GPU optimization Deep Learning with PyTorch – training and deployment of custom models 3D Reconstruction – NeRF, Gaussian Splatting, SfM, pose estimation Visual SLAM, Structure from Motion, Pose Estimation Computer Vision – OpenCV, feature extraction, sensor fusion ROS2, C++, Python Linux Systems, Shell Scripting, AWS S3 Performance profiling and memory optimization for GPU workloads If you’re looking for someone who understands both the science behind the model and the engineering required to make it run fast and efficiently on GPUs, I’d be glad to collaborate.