AI Multi-Agent Resume Analysis and Career Intelligence System

Author

A. Kanyaga, D. Natchatra, K. Prathipa, Mr. Varun Kumar. B


Abstract

Job seekers often struggle to understand why their resumes are rejected, which roles genuinely suit their skills, and what they should learn next. Applicant tracking systems (ATS) and recruiters screen resumes quickly, and most candidates receive little structured feedback. This paper presents the AI Multi-Agent Resume Analysis and Career Intelligence System, a web-based application in which a candidate uploads a resume, a pipeline of cooperating software agents extracts skills and experience, evaluates ATS readiness, collects and classifies live job listings, ranks them against the resume, identifies skill gaps, and prepares tailored application material and interview questions. The proposed system aims to combine secure registration and login, resume parsing, explainable resume analysis with severity-rated issues, job description matching, semantic job matching, skill gap analysis with a four-week learning roadmap, generation of a tailored resume, cover letter and HR email, an application tracker, and a consent-based recruiter connection feature. The application uses a Next.js and React frontend, a FastAPI backend, LangGraph for multi-agent orchestration, SQLAlchemy with SQLite or PostgreSQL for data storage, ChromaDB with sentence embeddings for semantic similarity, the Groq API for language-model tasks, and the Jooble API for job listings. Language-model output is treated as assistance, deterministic fallbacks keep the core analysis available when an external service is unavailable, and the generated documents use only facts present in the candidate's resume. The system is designed with consideration for usability, privacy, security, and reliable handling of uploaded resumes. Functional and integration testing can be used to evaluate the individual agents and their interaction. The proposed system provides a foundation for combining resume analysis, job matching, and career guidance in a single application, with further evaluation required on real-world resumes and job data.


Keywords

Multi-Agent System, Resume Analysis, ATS Score, Job Matching, Skill Gap Analysis, LangGraph, Large Language Model, FastAPI, Next.js



Full Text:

Download Paper PDF


References


  1. Ramírez. FastAPI Documentation. Available at: https://fastapi.tiangolo.com/
  2. LangGraph Documentation. Available at: https://langchain-ai.github.io/langgraph/
  3. GroqCloud API Documentation. Available at: https://console.groq.com/docs
  4. ChromaDB Documentation. Available at: https://docs.trychroma.com/
  5. Reimers and I. Gurevych, "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks," in Proc. EMNLP-IJCNLP, 2019, pp. 3982–3992.
  6. Jones, J. Bradley, and N. Sakimura, "JSON Web Token (JWT)," RFC 7519, Internet Engineering Task Force, 2015.
  7. Provos and D. Mazières, "A Future-Adaptable Password Scheme," in Proc. USENIX Annual Technical Conference, 1999.
  8. SQLAlchemy Documentation. Available at: https://docs.sqlalchemy.org/
  9. Next.js Documentation. Available at: https://nextjs.org/docs
  10. Jooble REST API Documentation. Available at: https://jooble.org/api/about
  11. OWASP Foundation. OWASP Application Security Guidance. Available at: https://owasp.org/

 

 

Share your valuable work from Social Media Buttons