Abstract: The goal of this paper is to develop a parallel algorithm that, on input of a learning sample, identifies a regular language by means of a nondeterministic finite automaton (NFA). A sample is a pair of finite sets containing positive and negative examples. Given a sample, the minimal NFA that represents the target regular language is sought. We define the task of finding the NFA, which accepts all positive examples and rejects all negative ones, as a constraint satisfaction problem, and then propose a parallel algorithm to solve the problem. The results of preliminary computational experiments on the variety of hexapeptide test samples are reported.