Quantum field theory (QFT) is currently recognized as a foundational framework in modern physics.It is used to characterize and analyze a wide range of physical scenarios in particles physics, condensed matter physics, cosmology, and has influenced many areas of pure mathematics. A common difficulty that arises in QFT analyses is when the interactions become strongly coupled. The physics can often be illuminated by understanding the symmetries and ’t Hooft anomalies of the system since these are preserved under renormalization group flow. When the QFT is supersymmetric, another powerful technique is to embed the QFT inside of a string theory construction whereby the internal geometric details of the extra dimensions in string theory can translate into the strongly-coupled dynamics of the QFT. The work in this thesis lies at the intersection of these two powerful techniques. In Part I, we discuss several new geometric constructions of 4D and 3D QFTs from M-theory on G2 and Spin(7) spaces with and without T-branes. We also discuss the conditions by which a 3D N = 1 theory engineered from M-theory on our Spin(7) backgrounds preserve spatial reflections and compute their ’t Hooft anomalies in a top-down fashion. In Part II, we introduce new techniques to extract the symmetries of various QFTs engineered from various string theory setups. Together, these works take into account both the charged operators and the topological symmetry operators in that these field theory objects are each given explicit string theory interpretations. Additionally, we utilize the conjectured absence of global symmetries in quantum gravities, to argue for a new, non-supersymmetric 7-brane in IIB string theory which we refer to as the R7 brane.