In this paper effort has been made to capture the materials behaviour into the Artificial Neural Network models, which can help to manufacturing industries for designing the processes like forging, rolling and extrusion etc. The materials data from the variety of alloy groups have been used to train the model and these are unique data since those are validated in the industry shop floors. The first model captures the pattern to identify the Dynamic recrystallisation (good) domain features, in which the actual process will have better microstructural control. The second model captures the undesirable regimes, where the instability occurs, this is the region has to be avoided during the material deformation. It is shown that both the models can be used on the newer material to get better results. These models can be used as an alternate to the dynamic materials model and as well as the expert’s interpretation skills to extract the process parameters for good and bad domains.