The article discusses a new way to estimate costs for road projects in Ethiopia using neural networks. The researchers created a model that can predict costs with an average error of 32.58%, based on only 48 examples. They believe that with more data, this model could help financiers, employers, and consultants make more accurate estimates. They also developed a user-friendly interface to make it easy to use the model. Overall, the study shows that neural networks could be a useful tool for estimating costs early on in Ethiopian road projects.