We present highlights of NCSA Gravity Group's trans-disciplinary program to fuse the computational power of the Blue Waters supercomputer with innovative applications of machine and deep learning to create scenarios for multi-messenger astronomy. We describe the construction of a novel computational framework that combined for the first time the Blue Waters supercomputer, containers and the Open Science Grid to validate the first detection of colliding neutron stars in gravitational waves and light; the use of numerical relativity simulations to model the formation and merger of black hole binaries in dense stellar environments; and pioneering work at the interface of HPC and artificial intelligence to disrupt the paradigm of gravitational wave data analysis. Finally, we present a new open source package to monitor and post-process large-scale numerical relativity campaigns to validate the astrophysical origin of gravitational wave discoveries with Blue Waters.
