Multiple quality characteristics are a serious concern for the rapidly increasing sector in recent years. To address this issue, the current study describes a novel approach to optimizing several rollers burnishing quality characteristics called as the desire function. Various multi-character models have been constructed by adjusting roller burnishing settings between lower and upper limits for different combinations of responses with the goal of improving surface roughness. Number of tool passes, spindle speed, feed rate, and interference were used as model variables, whereas surface roughness, dimensional deviation, cylindricity, and parallelism were used as response features to construct prognostic models. Practical work was carried out on the Al alloy 6061 using a 5-level design. The results showed that the greatest desirability of 0.943 was attained for all four response characteristics at an ideal spindle speed of 50.001 rpm, an interference of 5.747 mm, a feed rate of 0.095 mm/rev, and five tool passes. The computed model’s validity at maximum desirability value has been confirmed by executing confirmation experiments in the most favorable scenario. The desirability technique is effective for optimizing multi-response characteristics throughout the roller burnishing process, as demonstrated by the results of optimum process parameter combinations.