Artificial intelligence system for real-time training and calling and processing method thereof

摘要

The present invention relates to an artificial intelligence (AI) system for real-time training and calling, which is capable of efficiently interlocking a heterogeneous legacy system with an AI server, and a processing method thereof. According to one embodiment of the present invention, the AI system for real-time training and calling comprises: an interface middleware device receiving data from at least one legacy system to convert the received data into a data type to be processed in the AI server; a data handling device storing the data converted in the interface middleware device for each legacy system; the AI server requesting the converted data to the data handling device to train AI for the at least one legacy system; and an AI service device using information about the legacy substrate to be trained, information about data to be trained and AI training model information to handle AI training for the at least one legacy system and handling a training result in an AI training model performing the AI training therein.