Personality trait classification from natural language has become a significant area of research, driven by applications in psychology, human resources, and personalized marketing. Traditional methods often rely on structured questionnaires and psychological assessments, which, while effective, can be time-consuming and limited in scope. The advent of Natural Language Processing (NLP) and machine learning, particularly with the development of Large Language Models (LLMs), offers new avenues for automatically extracting personality traits from unstructured text data. This paper explores the application of LLMs in personality trait classification, particularly focusing on the Big Five personality model, a widely accepted framework in psychological research. We propose an advanced system that integrates data augmentation techniques and a guided reflective dialogue approach using Retrieval-Augmented Generation (RAG) to enhance the input data quality and quantity. Our system aims to improve the accuracy and reliability of personality trait detection by leveraging the strengths of LLMs and state-of-the-art NLP techniques.