The Enterprise Resource Planning framework (ERP) has been called attention to as special data frameworks paradigm. In any case, accomplishing an appropriate degree of ERP achievement depends on an assortment of factors that are identified with an association or venture condition.Those ERP projects should be satisfied by the customers and vendors in terms of ease of accessibility, flexibility, efficiency and reliability. In our existing work, AHP-RCF method uses requirements in the rank based priority level. However, it has not been discussed about the decision criteria of the customers. In our paper ,a mix approach between the AHP and ANN has been created to assess and choose the good degree of customization that is the requirements that can perform customization in a well efficient manner. The proposed method HAHP-ANN is used to measure the weight of customization and various structures of multi-layer neural networks have been analysed for the optimization. Also the learning of projectinformation gathering is done by using mix SVM classification approach based on which dynamic updation about the projectneeds can be provided to the user customization. The general valuation of the research strategy is led in the java modelbackgroundcondition from which it is demonstrated that the proposed research procedure prompts give the ideal result than the current research systems