The enhanced variation of the magnetic field during severe to extreme geomagnetic storms induces a large geoelectric field in the subsurface. Grounded infrastructure can be susceptible to geomagnetically induced currents (GICs) during these events. Modeling the effect in real-time and forecasting the magnitude of GICs are important for allowing operators of critical infrastructure to make informed decisions on potential impacts. As part of the UK-funded Space Weather Innovation, Measurement, Modeling and Risk (SWIMMR) program, we implemented nine research-level models into operational codes capable of running consistently and robustly to produce estimates of GICs in the Great Britain high voltage power transmission network, the high pressure gas pipeline network and the railway network. To improve magnetic coverage and geoelectric field modeling accuracy, three new variometer sites were installed in the UK and a 3 year campaign of magnetotelluric measurements at 53 sites was undertaken. The models rely on real-time ground observatory data and solar wind data from satellites at the L1 Lagrange point. A mixture of empirical machine learning and numerical magnetohydrodynamic models are used for forecasting. In addition to nowcast capabilities, contextual information on the likelihood of substorms, sudden commencements and large rates of change of the magnetic field were developed. The final nowcast and forecast codes were implemented in a cloud-based environment using modern software tools and practices. We describe the process to move from research to operations (R2O).