In this paper, we propose an object recognition platform based on artificial intelligence system which helps users to easily implement their smart applications with web user interface. The proposed platform provides object recognition functions based on Convolutional Neural Network system. The platform has a feature that it can be utilized through web browser without additional software installations since it has been implemented with some javascript libraries. Our platform consists of three modules which performs extraction of objects from input images, learning specific objects (such as a thing or face) from training dataset, and decision with inference based on prior learning, respectively. With the proposed platform, a user can create and manage a project for his or her own work, which operates learning and decision under one of two modes, face or object recognition, according to the user’s choice. Our platform also provides an interface to deliver the decision results to or directly interact with another system. This would enable our platform to be widely grafted to various application services based on Internet of Things.