Recruitment has become a lot more complicated with a surge in the number of applicants for each job. Traditional resume screening methods are time-consuming and subjective-and that leads to inefficiencies and biases in candidate selection. What we need is a more efficient, fair and accurate way to evaluate resumes. That’s where a new resume screening system comes in. By using Retrieval-Augmented Generation (RAG) with the Gemini API, we can improve the accuracy, efficiency and fairness of resume evaluation. That system pulls out the most relevant information from resumes, compares a candidate’s skills to the job description and gives you a selection probability score-complete with reasoning. We combined Natural Language Processing (NLP), machine learning and a resume dataset from Kaggle to refine the evaluation process. The experiments show the model can identify missing skills, match candidate qualifications to job requirements and give job seekers valuable insights. Integrating RAG into the hiring process can make a real difference to recruitment outcomes-and reduce bias and manual workload.