Application Development Using Large Language Models

This thesis focuses on learning methods and tools for building applications using large language models with the ultimate goal of developing a chatbot that can answer questions on a specific topic. As part of my thesis topic, the chosen topic was the development of a chatbot using language models to automate and simplify the processes of obtaining information about admission, study programs and university life for people interested in enrolling in the Higher School of Economics. To implement the task, the following technologies were used: LangChain for integrating language models, a vector database for storing data, ReAct Agents for planning tasks and selecting tools, RAG for supplementing knowledge about language models with new data. and LangSmith for monitoring and testing the system. The result was the creation of a functional and effective tool to improve information support for applicants.