Lawrence Livermore National Lab (LLNL) uses supercomputers to perform complex physics simulations. Maintaining parallel code is difficult when faster computers with different architectures are installed every few years. Portability layers can simplify this code by hiding details of computer architecture and parallelism. The goal of the clinic team was to improve RAJA, a portability layer created and used by LLNL. The team also com-pared RAJA to other portability layers with respect to usability and performance.