This paper is part of a two paper series on optimization methods for film cooling which seek to address the limitations of experimental optimization by utilizing advances in RANS based CFD to quickly optimize film cooling hole geometries. In the companion paper  on parametric optimization the optimum hole was experimentally demonstrated to have > 40% improvement in spatially averaged effectiveness compared to a baseline 7-7-7 hole, and was developed by leveraging RANS as a proxy for experimental data. In this paper adjoint based optimization was used to develop unique film cooling hole geometries. Adjoint optimization moves beyond using RANS as a proxy for experimental data instead utilizing the derivatives available in RANS to fully optimize the geometry of a shaped film cooling hole. The resulting geometry was experimentally validated to further increase performance by over 80% compared to the baseline 7-7-7 shaped hole. The study also show that further increases in performance are predicted when expanding the optimization target region. Furthermore, these new optimized geometries are readily manufactured by Additive Manufacturing (AM) processes and significantly less time consuming to generate than an equivalent parametrically optimized hole shape. These methods provide the tools necessary to fully utilize the large design space offered by AM and will be dramatically shift the future of film cooling hole design.