Predicting investment performance is central to attracting investors in any property or business venture. Investors are keen to predict the future in order to protect their investments and choose assets with the best returns. All asset returns are predictable to some extent, with returns on real estate relatively easier to forecast due to the nature of assets. Forecasting is thus an important component in property investment decision-making. Currently, majority of investors in Kenyan real estate sector, rely on speculation and sales comparison to make investment decisions. Multiple regression models have been applied successfully in forecasting real estate investments in other markets. They incorporate socio economic variables, housing and proximity characteristics to estimate the value of real estate assets. The researcher applied a multiple regression model for predicting house prices by setting house price as the dependent variable (Y) while holding the Gross Domestic Product, income of households, cost of land and housing units developed as the predictor variables (X).This predicted house prices (Y) on the basis of the X variables and determined the influence of the variables on the price. Agile development methodology was applied in the development a web application that integrated the forecasting model, an analytical backend helps to present the forecasts to investors in terms of figures, charts, and graphs that are easy to interpret and compare. Various tests were also performed on the prototype including integration and system tests. User acceptance testing was also carried out where majority of the respondents found the interface easy to use, and indicated that the application met its stated objectives as outlined in the usability questionnaire