Over the past few decades various quantum technologies have been proposed and it is their potential impact on the world that is making many researchers eagerly pursue their realization. This endeavor comes with new challenges that we do not encounter in classical computers for instance, as a fundamentally different set of principles determined by quantum mechanics governs the behavior of these novel systems. Quantum information is very delicate and prone to distortion by the environment, therefore the control of such devices needs to be carefully designed with the underlying physical properties in mind. Furthermore, in the case of quantum computing, algorithms and compilation techniques can make use of hardware noise characteristics to improve the quality of results. While eventually we would like to build a computing stack where these high-level design choices can be agnostic to such low-level details, this will not be possible in the near future as complete fault-tolerance is a major research goal that the community regards as a long-term mission. In this thesis we explore how the low-level understanding of a quantum system aids the optimization of different layers of its device stack. In particular, one major contribution lies in studying the design, improvement and implementation of control pulses that make a system undergo desired operations --- in the case of quantum computers (our main focus), this corresponds to the realization of gates that compose programs. However, we further explore quantum sensors as another technology that is promising to have practical use. Here control pulses determine the sensing scheme and the sensor sensitivity, which one seeks to maximize. Pulse optimization covers a lower layer in the quantum stack we study first. We then turn to the top layer and discuss work that improves a particular quantum algorithm, the Variational Quantum Eigensolver, by including efficient hardware noise modeling in its execution. Overall, we demonstrate the impact of design choices for quantum technologies that are inspired by the physics governing the underlying systems.