The problem of the house sales price inconsistency and its influence on the Socio-economy is more complex in developing countries that need organized house price prediction methods. The study aims to develop a house price prediction model for all stakeholders who have a stake in the selling and buying process to make an effective decision. In the study, necessary procedures have been done to realize the house price prediction for the best decision making. Design science research methodology was applied and the study considers all documented house sales transactions amounted to 11565 sales data of Bahir dar town of the last decade as a sample of the study which was analyzed through python. 70% of the data were used for training and 30% of the dataset were used for testing. ANN, support vector machine, gradient boosting, random forest, linear regression, and decision tree regression. These algorithms are trained and tested on a labeled dataset and run with this data set through a series of pre-processing steps. Based on the finding of the study sold year, building cost, kebele, location, and area of the land are found to be the most valuable features of a house for predicting the housing market. The study investigated selecting the appropriate machine learning techniques for predicting house price and developing the model with appropriate data and selecting the fitted model. The finding of this study proves random forest regressor is the best among trained and tested predicting models gives an accuracy of 89% which is taken as an appropriate model to predict house price in Bahir dar town. Based on this result, a prototype of the System has been developed and tested which is sound to predict house prices in Bahir dar town.