MatchScore — AI Agent for CV ↔ Job Description Matching

MatchScore is an AI agent that compares a candidate’s CV with a Job Description and produces:

• Match Score (0–100)
• Strengths & gaps
• Actionable recommendations for improving fit

Demo

• 🎥 Video demo (Google Drive): https://drive.google.com/file/d/1lQfXs_NPcM24vrzjNgJuI0ELSf5v9kkt/view?usp=drive_link
• 🤖 Agent link (ChatGPT): https://chatgpt.com/g/g-6973b0c982ec8191a540f05cd1d20737-matchscore

What it does

1. Extracts structured requirements from the Job Description (skills, experience, tools, must-haves).
2. Parses the CV into a structured profile (skills, roles, achievements, tools).
3. Compares both sides and assigns a transparent score.
4. Returns:
   • Matching highlights
   • Missing requirements
   • Suggestions to close gaps

Key Features

• ✅ Structured scoring (consistent evaluation)
• ✅ Gap analysis (what’s missing + why it matters)
• ✅ Explainability (clear reasoning + evidence-based output where applicable)
• ✅ Responsible AI considerations (privacy + bias awareness)

Tech & Approach

• Prompt engineering for structured extraction + evaluation
• Step-by-step decomposition (requirements → profile → scoring → recommendations)
• Output formatting for recruiter-friendly summaries

Repository Structure (recommended)

• README.md — project overview (this file)
• assets/ — screenshots / GIFs
• docs/ — presentation / writeups
• examples/ — anonymized sample inputs & outputs

Privacy Notice

Do not upload real CVs with personal details.
Use anonymized examples (remove phone, email, addresses, IDs).

License

Choose one if you want (e.g., MIT). If this is a course project, you can also keep it “All rights reserved”. This project demonstrates a general-purpose scoring & recommendation pipeline that can be adapted to matchmaking, player profiling, or content recommendations.