By their very name, defects in crystalline materials are often assumed to be deleterious. While this is often accurate, defects also enable a wide variety of properties not observed in the hypothetical, perfect state. Studying the control and impact of defects is not trivial - defects are typically dilute and their concentrations depend strongly on subtle changes in synthetic conditions. Fortunately, the last decade has seen significant growth in our ability to (a) accurately predict defect concentrations as a function of synthetic conditions and (b) estimate the impact defects have on material properties. In this talk, we will begin with a comprehensive introduction to modern computational defect theory and the underlying thermodynamics of rational synthetic design. With this foundation, we will explore case examples from within our NSF DMREF (2017-2023) to unite theory and experiment in the control of defects for thermoelectric applications. This team brought together expertise in defect calculations (Stevanovic, CSM; Ertekin, UIUC), thermoelectric materials synthesis and characterization (Toberer, CSM; Snyder, NW), and scattering (Toney, CU Boulder). Specific material systems will include thermoelectric materials Zn13Sb10, Mg3Sb2, CuInTe2, and Hg2GeTe4 as well as the kagome superconductor KV3Sb5. Particular emphasis will be on using defects for achieving desired functionalities (resonant levels, phonon scattering, electronic phase transitions, etc) rather than simply reducing their concentration.