Energy consumption in Office buildings (OB) can be significantly reduced with suitable façade design solutions. In order to achieve a highly effective solution in the field of energy efficiency, many complex procedures are required in the process of simulation and equation building. This gives rise to the necessity of creating a prediction model of the influence of façade types on the Energy Use Intensity (EUI) within office buildings (OB). In this study, a framework of a machine learning model is made to forecast the energy consumption of OBs based on the influence of façade features. OB is influenced by 3 input variables (Window to Walls ratio-WWR, façade north offset, shading divide length) affect two output variables, namely Energy Used Intensity (EUI) and Carbon Dioxide Emissions (CO2). Data collection was conducted through energy simulation based on computer-aid design software (CAD) called Rhinoceros and grasshopper. The energy intensity simulation tool is ClimateStudio-Solemma, a highly reliable tool in many studies of energy simulation in the building sector. The machine learning model implemented during EUI and CO2 prediction is the Adaptive Neuro Fuzzy Inference System (ANFIS). The results show that the effects of the passive system significantly affect energy consumption inside the building, and the EUI and CO2 Emission prediction models demonstrate very low RMSE and R2 results, opening the potential to replace traditional simulation and calculation models, which require a lot of time and technical expertise.