Across the millimeter and submillimeter sky, we see relic radiation giving evidence that the early universe went through a hot, dense phase out of which all modern structures (stars, galaxies, galaxy clusters, etc.) eventually coalesced. Precision measurements of this light, the cosmic microwave background, serve as a foundation to our modern understanding of the universe and present us with a means to probe physics at a wide range of scales. Higher precision measurements will further inform our models of the universe and provide constraints on beyond-standard-model physics, such as dark energy, inflation, and the existence of additional light relativistic particles in the early universe. Attaining these higher constraints with upcoming ground-based observatories such as the Fred Young Submillimeter Telescope and the Simons Observatory will require unprecedented numbers of superconducting detectors operating at nearly the fundamental limits of sensitivity. One such detector technology is the kinetic inductance detector (KID), a superconducting resonator that allows for natural multiplexing and photon-limited performance. Prime-Cam, one of two primary survey instruments for the CCAT collaboration's six-meter Fred Young Submillimeter Telescope, will ultimately deploy more than 100,000 KIDs across seven independent instrument modules. In this thesis, we present an overview of some of the author’s contributions to the field of experimental cosmology as a member of the Atacama Cosmology Telescope, Simons Observatory, and CCAT collaborations. In particular, we describe the development, design, and test of many key elements of the detectors and readout for the 280 GHz and 350 GHz instrument modules for CCAT's Prime-Cam receiver. We provide a comparative analysis of aluminum and titanium-nitride-based KIDs, the two most common KID materials at millimeter and submillimeter wavelengths, both of which are being used in Prime-Cam. We then describe the cryogenic readout system for the 280 GHz instrument module, including a demonstration of photon-noise limited performance with prototype detectors. Next, we detail the cryogenic focal planes and detector array modules for the 280 GHz instrument module and provide status updates on the three completed arrays. We conclude with a discussion of several interesting science cases that these technologies may enable when deployed.