Quantum Computing is in an era of growing pains. As quantum algorithms become more tuned to the computational capabilities of near-term quantum devices, compilation techniques that fully utilize the architecture while avoiding potential pitfalls are a necessity. Many different types of quantum devices are being developed each with its own advantages and unique features. It is unclear which architecture will be dominant, and many are in active development. Some problems, such as limited connectivity, imperfect gate execution, and lower coherence times are shared across architectures, and mapping, routing and communication algorithms have been developed to circumvent and avoid these issues. But, each architecture has its own set of specific problems to overcome. It is not simply the case that a single compilation pipeline or set of algorithms will be able to make the best use of each of these architectures. This work proposes an exploration into adapting a general quantum compiler algorithm to several different architectures. This thesis will focus on two main architectures, Neutral Atom and Superconducting qudit devices. Neutral Atom architectures have additional flexibility in routing qubits beyond nearest neighbor connectivity, while incurring some serialization cost due to increasingly large ``areas of restriction''. But, we are still able to take advantage of these more flexible interactions through careful mapping and routing. Additionally, we can craft strategies that make use of these features to circumvent Neutral Atom Architectures' biggest execution downfall: atom loss. Many quantum architectures can be naturally extended to extra computational states beyond the traditional |0> and |1> used in classical computing and most quantum computation. These extra states can be used to reduce the number of operations, but at the cost of slower operations and decreased stability of the computational units. Once again, we can adapt an existing base algorithm to mix-and-match qubit-only, mixed-radix, and higher-radix operations to develop efficient programs that make use of less resources and increase the chances of successful quantum computation.