An effective method of fluid application such as high-pressure coolant (HPC) augments the performance characteristics by producing quality products. Effective control of parameters, prior to actual machining, prevents the loss of resources which in turn maximize the productivity. Thus, an adequate prediction model of surface roughness and an optimization model of control parameters must be determined that can be efficiently used for HPC employed machining. In this regard, this article presents the formulation of two predictive models of surface roughness, one by using artificial intelligence-based technique, namely support vector regression (SVR), and another by applying conventional technique called response surface methodology, in turning of hardened and tempered AISI 1060 steel in dry cutting and under the application of pressurized oil jet. The cutting speed, feed rate and material hardness were considered as input variables for model formulation, and based on these factors, the full factorial experimental design plan was conducted. The performance of the predictive models was evaluated on the basis of root mean square error. Additionally, the effects of control factors were evaluated by using analysis of variance. Furthermore, separate optimization models were created using composite desirability function and genetic algorithm (GA) to determine the control factor setting corresponding to minimal surface roughness. Both of the optimization models suggested an optimal parameter setting at 0.10 mm/rev feed rate, 161 m/min cutting speed and ~43 HRC material hardness. The adequacy of the optimization models was evaluated by a confirmation test. The predictive model by SVR and optimization model by GA provided the highest accuracy.