The aim of the project is to implement a universal prototype system for dialogue between man and computer in a call centers. The system of automatic recognition of speech distorted by the system of telephone communication will be developed. The system will consist of three modules: Polish automatic speech recognition, semantic analysis with the choice intelligent responses and speech synthesizer. The first module will be improved and adapted version of an automatic speech recognition system that was developed at AGH. Implementation of the second module will be based on WordNet and a semantic dictionary. The project will use the existing commercial solutions for speech synthesis.