In the large-scale production of Diethyl Oxalate (DEO), numerous control and optimization challenges are faced. These issues are targeted in this study by implementing a gain scheduling approach to control a catalytic fixed bed multi-tubular reactor designed for DEO production. A novel inferential modeling method, developed using Python and response surface methodology, was applied to simulate the dynamics of the reactor, and the inlet coolant temperature was identified as the most influential parameter. An inferential control system was implemented that utilized gain scheduling for Proportional Integral Derivative (PID) controllers. The results showed that the proposed PID controller gains offered fast and stable responses, even at high conversion setpoints. However, minimization of response overshoot at high amplitudes of step changes proved challenging, leading to a decrease in conversion estimator accuracy. The use of a quadratic model from response surface methodology for the regression of the inferential estimator was proposed, showing improved prediction accuracy over the linear model. These findings provide valuable insights and guide future optimization in the design and operation of reactors for large-scale DEO production.