A neural network model was created to estimate costs for highway projects using data from projects in Newfoundland, Canada. Instead of using complex software, a simple spreadsheet format was used for construction professionals. The model's weights were determined using optimization techniques to improve cost predictions. A user-friendly interface was developed to make it easy for users to input data and get cost estimates. The model can also adapt to new data and account for project uncertainties. This approach can help construction practitioners accurately estimate costs for highway projects.