We utilized Bayesian optimization to improve the thermoelectric performance of Sn-doped InGaAsSb thin films for flexible thermoelectric device applications. We explored a 5D space and obtained a figure of merit ZT = 0.033 at room temperature, which represent a 2.5-fold improvement compared to random initial experimental results. These findings not only promote the development of thermoelectric devices utilizing III-V semiconductors but also highlight the effectiveness of Bayesian optimization in multidisciplinary materials.