Depending on the type of the customer, there are some specific configurations about enterprise customers, for example, a view of different users of that enterprise are given in a searchable list. For each individual user, after registering and creating an account, there is a procedure for training the HPA internal classifiers with the biometric characteristics of that user. Depending on the types of the HPA inputs for which that user has registered, different number of repetitions may be necessary, in order to achieve a certain high level of precision and low level of unwanted login rejections.
During this implementation phase, the administrator may group the users and set different general thresholds to whole groups, as well as specific individual thresholds, thus allowing fine-tuning of the percentages and promilles of the false rejections of the system.