**Abstract:** This research proposes an automated system for optimizing affinity chromatography purification protocols using a combination of Reinforcement Learning (RL) and Bayesian Hyperparameter Optimization (BHO). Focusing on His-tag purification, a prevalent technique in protein biotechnology, the system dynamically adjusts key chromatographic parameters—flow rate, buffer composition (pH, salt concentration), and column loading—to maximize protein yield and purity while minimizing resin usage and processing time.