We developed advanced forecasting models to accurately predict energy cost, demand, generation, and carbon and renewable fuel mix across the council estate. This required detailed modelling of the council’s existing infrastructure, generation assets, and the inherent variability of renewable sources. The objective was not only to understand current capability, but to inform future infrastructure investment and support strategic planning around emerging demands such as EV charging. To achieve this, we built machine learning and AI models capable of predicting next-day costs and weather-driven variables. Data availability presented significant challenges, including reporting delays and inconsistencies in historical datasets. These issues were further compounded by Covid-19, where reduced building usage distorted consumption patterns. We addressed this through advanced data correction techniques, careful tuning of model hyperparameters, and iterative testing over a three-month period to achieve the highest possible predictive accuracy. The model integrates live condition data from OikoLab for weather metrics, SEMO for electricity pricing and volumes, and EirGrid for fuel mix modelling. Solar generation was forecast using a Random Forest model trained on council-specific generation data. Biogas, wind turbines, and solar assets were all factored into a unified view across 22 council buildings.