Jamie Gooding developed Gemm autodiff pullback support within the ferdymercury/root repository, focusing on enabling differentiable Gemm workflows for TMVA SOFIE. By implementing a custom Gemm_Call_pullback in C++, Jamie addressed the backward pass of the Gemm operation, which is essential for automatic differentiation in machine learning models. The work included creating TestGemmDerivative.cxx to rigorously verify the correctness of the derivative computations, ensuring robust test coverage and reducing the risk of regressions. Jamie’s contribution demonstrated depth in linear algebra, automatic differentiation, and testing, resulting in a maintainable and reliable foundation for gradient-based optimization in TMVA SOFIE.