This thesis explores Deep Reinforcement Learning (DRL) strategies to achieve effective gust alleviation on a lifting surface. Experiments were conducted using a NACA 0018 airfoil equipped with a flap to dynamically modulate the pressure distribution on the airfoil in response to impending gusts. The experimental setup integrates five pressure sensors distributed along the chord and a hot-wire anemometer in the wake to monitor in real-time the flow conditions. The flap displacements, effectively corresponding to the control strategy, is governed by a Soft Actor-Critic (SAC) algorithm implemented within a custom Python environment. An agent is trained to attenuate the longitudinal velocity fluctuations in the wake.