Traditional resume screening methods, which are time-consuming and prone to inconsistency and bias (due to human evaluators or keyword-based ATS), fail to understand the contextual meaning of candidate requirements. The purpose of this paper is to develop an Explainable and Bias-Aware Large Language Model (LLM)-Based Resume Screening System for evaluating candidates' qualifications and achieving better matching by incorporating NLU approaches with LLaMA 3 to understand candidate resumes and job descriptions contextually rather than relying on keyword-based similarities. This system will score candidates' skills, qualifications and experience with generation of explainable output, and the bias awareness mechanism is added to ensure fairness and transparency during recruitment processes. This system incorporates the functions of resume preprocessing, candidate job matching by semantically matching resumes with job descriptions, candidate scoring, explanation generation and candidate scoring display, and the system is implemented with an interactive Streamlit application interface. Experiments results shows an enhancement in contextual understanding and more accurate candidate-job matching.