Recently, sustainability is becoming a more important aspect in the building sector. The use of performance-driven approaches are progressively growing interest among architects and engineers. Proper design of a large span roof shading system can positively influence the micro-climate of the space underneath, representing a significant source of daylight and reducing the building energy demands. To support the decision making process and steer towards high-performing solutions, the adoption of computational design processes has the potential of being a fast and reliable approach. This project was undertaken to propose a computational workflow and evaluate its effectiveness as supportive decision-making tool from the early design stage of large roof shading systems. The proposed Computational Design Exploration (CDE) is adopted to evaluate three different concept alternatives in terms of daylight and thermal performances. Based on the visualization and analyses of the data, the best performing morphological features of the three alternatives are identified and the design is refined to create an optimal fourth concept. In a second phase, the Computational Design Optimization (CDO) workflow is applied to obtain high performing solutions in terms of daylight objectives, by varying geometrical and material inputs. 50 configurations are selected as optimal. The post process of the CDO consists of a second CDE, in which the selected samples are evaluated in terms of energy performance. The final shading system is chosen by identifying the input settings that allow the lowest energy consumption. As result, the daylight requirements are fulfilled and the thermal properties are used as final decision criteria. The steps described in the CDE and CDO workflow are followed through the combined use of a parametric modelling tool (Grasshopper) and a multidisciplinary design optimization platform (modeFRONTIER). Post-processing tools are adopted to help the identification of interaction effects of the variables on the performance targets. By a series of data visualizations and sensitivity analyses, it is determined whether the final optimal selected design configurations improve the visual and thermal performances. The most important identified trends and variables are used as main inputs to produce a more structured and integrated computational workflow. The final proposed workflow is meant to be a versatile method to assist the designer in the decision making process, yet maintaining his/her autonomy of judgement. Future research is needed to prove the validity and effectiveness of the proposed workflow. The integration of other design variables and performances objectives could lead to an holistic method, suitable for every kind of multi-objective design problem.