This paper explores the basic concepts of connecting devices using intelligent services in the Internet of Things (IoT) space. It defines the concepts of objects, sensors, and actuators, as well as the concept of context awareness, which is crucial in the cognitive aspect of IoT services. Some existing solutions in the field are presented, including Apple Siri, Amazon Alexa, and SmartThings. The implementation of the SmartHome web application is described, which provides users with the ability to control connected air conditioners and lights. Control involves turning specific devices on and off, as well as adjusting temperature and fan speed for air conditioners, and changing light intensity for light devices. The application also offers a cognitive service that monitors user inputs for individual devices belonging to specific users and it automatically predicts device states based on these inputs and specific contexts. The tools used in implementing these functionalities are outlined: React, Typescript, and Bootstrap for the client-side, and Django for the server-side. Basic concepts of neural networks are presented, and the implementation of a neural network in Keras is described. A case study depticting everyday usage of the application with data from a hypothetical user is provided, demonstrating all the steps including configuring the application and showcasing the operation of the neural network predictions. The paper is commented on as a whole, and ideas for future improvements are presented.