Selective Laser Melting has been increasingly applied in industrial manufacturing due to a high density of printed products and more free design. However, a jungle of factors affects processes such as machine characteristics and materials properties. Moreover, ranges of crucial factors such as laser power, laser scanning speed, layer thickness hatch distance are wide. Therefore, it is difficult and requires much testing time and cost to select a suitable process parameter for manufacturing a desirable product. An Artificial Neural Network was applied to build an optimization system for finding out optimal process parameters. Inputs of the system are desirable properties of a product such as relative density ratio, and surface roughness while outputs are laser power, laser velocity, hatch distance and layer thickness. Applying the system not only requires less pre-manufacturing expenditure but also helps the printing users to choose approximate process parameters for printing out a desirable product.