Rahman does large-scale atomistic modelling of semiconductor quantum devices: tight-binding and DFT calculations of donor and quantum-dot wavefunctions, valley physics, spin-orbit coupling, hyperfine interactions and the response of all of these to strain and electric field, at system sizes large enough to represent a real device. The group works hand-in-glove with the Morello, Dzurak, Simmons and Rogge experiments, and increasingly uses machine learning to invert measurements into structural information. Positioned against the established body of NV-ensemble quantum sensing work — DEER, nanoscale NMR and T1 relaxometry protocols operating at pT/sqrt(Hz) field sensitivity — the same first-principles machinery is what predicts the hyperfine and spin-bath environment that determines T2 — and therefore the achievable pT/sqrt(Hz) sensitivity — of any solid-state spin sensor, including NV. Computational PI; would suit a candidate wanting a theory/experiment bridge role.