Personal morning brief powered by Large Language Models

Abstract

In this thesis, we embarked on a journey to address the challenge of information overload on the internet by developing a personalized morning brief website powered LLMs. Our objective was to streamline the information-gathering process, offering users a tailored experience that aligns with their individual interests and preferences. To achieve that, a backend service developed using FastAPI framework and a frontend service developed using React were created. Also, LLMs play an important role in our project, so we use ChatGPT-3.5 Turbo to empower the main functionality of our project.