Published Paper Details:

Generative Artificial Intelligence (AI), a subset of AI that generates new content such as text, images, and audio, has recently made significant advances, with key developments including Large Language Models (LLMs) such as OpenAI's Generative Pre-trained Transformer (GPT) -3/4, Large Language Model Meta AI (LLaMA), and Pathways Language Mode. These models, trained on large datasets to detect patterns and context inside language, have proven beneficial for difficult jobs for humans, providing efficiency, accuracy, and uniqueness. This research project uses OpenAI's GPT model to provide personalised applications like as document summarization, question answering, and numerical data analysis. Users may submit documents in a variety of formats and obtain a summary, as well as ask questions to extract responses. This function has several uses, ranging from students summarising educational materials to attorneys summarising court data. Users may also run data analysis procedures on CSV files containing numerical data, which is important for summarising financial information. As technology advances, the potential applications of generative AI models grow, influencing industries such as entertainment and healthcare. These models' versatility makes them useful tools for problem solvers in a variety of sectors. The fast development of huge language models and generative AI in recent years has been astonishing, with OpenAI's release of ChatGPT in 2022 pushing these technologies to the forefront. This paper studies the characteristics and applications of big language models, generative AI, and the features of Retrieval Augmented Generation (RAG). A software programme capable of interactive conversations in the context of specific papers is developed and tested, resulting in an application that is equivalent to commercial solutions. This programme can summarise large papers into simple, succinct English and provide accurate answers to proposed queries. Future optimisation efforts can concentrate on reducing reaction time and comparing the utilisation of various language models.