David Schmidig

David Schmidig

ML / CUDA Engineer


[about]




About Me


I am a Senior AI/ML and CUDA Engineer with a deep passion for creating cutting-edge solutions in computer vision, real-time AI systems, and scalable deep learning applications. With a strong foundation in advanced algorithm design and optimization, I have honed my expertise in bridging state-of-the-art research with practical, production-ready systems. My career is defined by a drive to solve complex challenges while fostering innovation and collaboration within diverse teams.



Over the years, I have successfully led cross-functional teams, guiding the development of sub-second human skeletal pose tracking systems, multi-modal human skeleton pose optimization, and action recognition pipelines. These projects required not only strong technical expertise but also a meticulous focus on building scalable, maintainable, and high-performance software. I take pride in delivering robust solutions that integrate seamlessly into production environments, ensuring reliability, efficiency, and long-term value.



With over 10 years of experience in C++ and Python, I am proficient in a wide range of tools and libraries, including high-performance computing frameworks (OpenMP, BLAS, Eigen, Ceres, Boost, TBB), GPGPU programming (CUDA, CuDNN), and deep learning frameworks (PyTorch, OpenCV, numpy, pandas, matplotlib/seaborn, scipy). My expertise extends to orchestration tools such as Union/Flyte and MLflow, which I leverage to streamline development, enable reproducibility, and ensure scalability in AI workflows.




My Motivation


At the core of my work is a passion for transforming state-of-the-art research into tangible, high-quality products. I am inspired by the process of unraveling complex problems and translating innovative ideas into systems that deliver measurable impact. My ultimate goal is to bridge the gap between academic excellence and industrial applicability, ensuring that the solutions I help create are both cutting-edge and practical.