This paper presents a time-optimized process planning system for mold/die polishing using various sizes of abrasive stones on a flat mold or a die surface. The system is assigned a specific set of abrasive grain sizes. User inputs to this system are measured initial surface roughness of the EDMed mold or die and the desired final polished surface roughness. The outputs of this system, optimized for minimum polishing time, are a selected sequence of grain sizes from among the available set with each grain size used for one polishing step. With each polishing step is an associated polishing time, polishing pressure and workpiece RPM. A series of initial polishing experiments were conducted for the different available grain sizes and workpiece roughnesses at different pressures and RPM with the change in roughness measured for polishing duration. From this, an expert database is constructed for designing the fuzzy logic rules. Then a neuro-fuzzy technique combining the learning ability of neural network and the inferring ability of fuzzy logic system is used to construct the exact membership function of the fuzzy interface. Finally,' actual experimental results from suggested polishing sequences are compared with the predicted value validating the proposed approach.