Workover rigs are a crucial resource in the exploration and production of oil and gas, being used in the wells maintenance operations. To avoid resource idleness and scarcity, oil companies plan which rigs will serve the wells and when the activities will occur in the so-called Workover Rig Scheduling Problem (WRSP). This decision-making emerges in a highly uncertain environment and the majority of literature approaches are deterministic models and heuristics. Aiming to assist the WRSP, this study proposes two data-driven joint chance-constrained (JCC) optimization models considering the uncertainty in the jobs processing times and the solution feasibility: a deterministic equivalent using mixed-integer non-linear programming and a scenario-based approach with mixed-integer linear programming. Chance-constrained models consider the risk in the constraints, generating solutions more adherent to reality. Our strategies focus on JCC models with right-hand side uncertainty. In turn, data-driven optimization is a new trend tackling the data-related uncertainty with machine learning and data science. Our data-driven approach uses clustering and data-mining to clear and retrieve information from the data and a Ridge Regression that predicts the endogenous uncertainties in the model. The regression uncertainty is inserted in the model. The next steps of the studies are to implement the Wasserstein distances method to reduce the number of scenarios and achieve practical results.