The increasing escalation of digital recruitment platforms have been growing, so the demand for efficient, automated systems and the need for effective, automated methods to match resumes with job descriptions (JDs) and eliminate the inefficiencies of human screening procedures have been increasing due to which the quick growth of digital recruitment platforms has paved the way for advanced application tracking systems. In order to extract and retrieve candidates resumes in accordance to skills, experience, position suitability, along with contextual relevance, advanced natural language processing NLP techniques have been utilized. This study presents a complex resume ranking system that makes use of cutting-edge natural language processing (NLP) and machine learning techniques. The model is a multi-level prototype focusing on advanced techniques like TextRank for getting extractive summarization from the resumes, Sentence-BERT (SBERT) for semantic similarity analysis where the meaning of the word is given much prominence, similarly DeBERTa for deep contextual understanding which focuses on what context the words have been discussed, and spaCy for named entity recognition are all integrated into the system. The prototype has been tested using ground truth data on multiple test cases, the system finds the best applicants for positions like Assistant professor, civil engineer, and software engineer with a high success rate. Regular expressions (regex) have been used to extract contact information and E-mail information, while SBERT and TextRank collaboratively provide brief summaries of candidate resume. The ranking system creates a composite score by combining cosine similarity for contextual overlap, Sequence Matcher for role alignment, difflib, and Jaccard similarity for skill matching. By generating the composite scores basing on the weighted sum of scores model, preference is given to the weights with their relevant importance. This method focuses on cutting labour cost during recruitment.