The objective of our work is to provide the automated planning of production processes necessary to manufacture mechanical products. This study of process planning activity pointed out three phases where the use of artificial intelligence techniques seems useful. In the first phase a module recognizes the features for design and converts them in features for manufacturing. The second phase is related to the tools selection and the cutting tool parameters definition for turning, milling and drilling operations; also herein we obtained the optimal result using the experience added to the background know-how. The last phase is related to operations schedule which evaluate several factors using the experience; this requires the creation of a knowledge-base, implemented by rules, and a research algorithm for the optimal solution. This paper examines in particular the second phase. Following to the analysis of the problem's domain we decided to implement the system on a personal computer, in MS-DOS environment, using Prolog II language. This choice has been suggested by the wide diffusion and the low cost of this platform. The expert system realized is composed of two parts: the tool selection and the cutting tool parameter determination. In the first part we implemented several rules for the definition of tool's features and a data base research to find the tools that satisfy those features. In the second part the cutting parameters are determined by a learning module based on the historical data. The system performs within good computational time; therefore its implementation in a production environment is realistic. The major advantages of our system are: a global approach to process planning and reduction of the interaction with the user, furthermore, our solution was verified by skilled experts performing a good results in tools selection and cutting parameters determination.