Forecasting is of extreme importance for companies as it is the input for the S and OP process, an essential part of Supply Chain Management (SCM); however, considering the close involvement of humans in various moments that compose this activity, cognitive biases (e.g., risk-seeking) and their influence represent a threat to an organization s performance, with many potential risks to supply chains. This dissertation establishes a novel framework for data-driven and cognitively aware supply chain management. Multiple forecasting models were evaluated using the proposed framework to select the most satisfactory model considering the organization s strategic vision. This allows the SCM manager to perform judgmental adjustments, which are evaluated through an automatic risk-seeking bias detection system. The dissertation was experimentally assessed over simulated scenarios from real data about cardboard production during 2017 and 2023. The results evidenced the effectiveness of the proposal in addressing the complexity and intertwined objectives of different stakeholders within a supply chain. The automatic definition of a preference-based rational model defined by the SC manager, made it possible to detect risk-seeking biases using different thresholds for judgmental adjustments, thus mitigating the adverse effects of risk-seeking biases. In summary, it can be argued that the proposed framework represents an important advance regarding the implementation of the Humachine paradigm, integrating the positive elements of advanced statistical modeling with human expertise and context.