Graham Markall is a software engineer in NVIDIA RAPIDS, where he leads work on Numba (a Python JIT compiler), supporting its use as a key part of the CUDA Python ecosystem. He has a background in compilers and numerical methods, beginning with his Ph.D. work at Imperial College that focused on runtime code generation for finite element solvers running on GPUs and multicore CPUs. Subsequently, he worked on compiler and math libraries for Python and Java, and on GCC- and LLVM-based toolchains for various specialized architectures before joining NVIDIA in 2019.