The take-off in Artificial Intelligence (AI) in the last decade has led to the development of such software tools as expert system shells, causing a tremendous incentive for applying AI in the Industry. In many cases, however, the original enthusiasm has gradually faded away as the results obtained did not meet the original expectations1 .
This paper deals with AI applications to welltest interpretation and presents what can be achieved practically today and what can be expected in the future. Most of the early work was focused on developing assistant programs, helping the interpretation engineer to select the appropriate solution model(s). The goal of this study is instead to facilitate and speed-up the task of the interpretation engineer. An interactive computer program developed in this framework is described and several real test examples are presented. The practical advantages and limitations of the program are described in detail.