This study presents the development and evaluation of a novel shell-and-tube heat exchanger (STHX) design with segmental porous baffles. Computational fluid dynamics (CFD) in combination with machine learning tools are utilized to investigate the thermo-hydraulic impacts of segmental porous baffles on shell side flow of a STHX. Three geometric parameters (number of baffles, baffle angle, and baffle thickness) of these baffles, which are placed inside the STHX, are selected to perform the parametric study and multi-objective optimization. Higher number of baffles are beneficial to increase the rate of heat exchange; however, it would escalate the pressure drop considerably. Results also show that baffles angle plays a critical role on the performance of a STHX. An artificial neural network (ANN) is trained to predict the system's performance. As lowering the pressure drop and increasing the heat transfer are the two main objectives in STHX, a multi-objective optimization study is conducted. Different decision-making algorithms are also applied to find the best alternative among the Pareto frontier points. Results of optimization show that a STHX with 10 porous baffles, baffle angle of 111.9, and baffle thickness of 16.69 mm would be the best geometrical configuration which results in a heat transfer rate of 523.81 kW while the pressure drop of the shell side flow would be 48.87 kPa. With this novel design, it is also possible to improve both pressure drop and heat transfer rate of STHX, simultaneously. A particular configuration of the introduced STHX could reduce the pressure drop by 61.3% while heat transfer enhances by 11.15% simultaneously.