Modern educational processes require automation to improve learning efficiency. The use of artificial intelligence (AI) allows to optimize the management of the learning process, increase the personalization of learning, and automate assessment. In the context of digitalization of education and the growing role of distance learning, it is important to create adaptive systems that meet the needs of students and teachers. Thus, the aim of the work is to develop and implement a web service that will support the learning process by automating the generation of test tasks, checking answers, and integrating with learning systems. A machine learning module has been implemented to automatically analyze student work (grading, checking for uniqueness). Natural language processing (NLP) was used to analyze student responses and create adaptive content. Automatic generation of test tasks based on learning materials is implemented, which increases the personalization of learning. Standard assessment systems (e.g., Moodle testing) are often limited to multiple choice, while the use of semantic analysis allows you to evaluate creative tasks and open-ended answers without manual verification by the teacher. Most LMS systems (Moodle, Google Classroom) provide standard content for all students, while the developed system adapts to the needs of a particular user, increasing the efficiency of learning. Instead of creating another isolated LMS system, the platform is designed with a flexible API that allows it to be easily integrated into existing educational solutions (for example, university portals).