nlp-resume-builder

---

📽️ Demo Video

▶️ Watch the YouTube demo (4 min)

---

Resume Scoring UI Setup

Instructions to Run

1. cd resume-builder-ui (move to the directory)
2. Install the dependency npm install
3. Start a development server npm run dev
4. Open your browser and visit http://localhost:3000 to see OpenResume live

---

Resume Scoring FastAPI Backend

This is a lightweight FastAPI service that scores a resume against a job description based on skill matching using TF-IDF weights.

🛠 Setup Instructions

1. Install required packages:

pip install fastapi uvicorn scikit-learn pandas spacy joblib
python -m spacy download en_core_web_sm

---

🚀 Running the API Server

Inside the fastapi-backend folder, run:

uvicorn main:app --reload

• Server will start at: http://127.0.0.1:8000
• You can access automatic API docs at: http://127.0.0.1:8000/docs

📨 Usage

Endpoint: /score_resume

• Method: POST
• Request Body (JSON):

{
  "resume_text": "<Paste your resume text here>",
  "job_description": "<Paste your job description text here>"
}

• Response (JSON):

{
  "normalized_score": 0.85,
  "matched_skills": {
    "java": 0.6,
    "javascript": 0.25
  },
  "missing_skills": {}
}

Endpoint: /bot/ping

• Method: GET

• Response (text):

true

Endpoint: /bot/chat

• Method: POST
• Request Body (JSON):

{
  "sender": "<User id>",
  "message": "<Paste your prompt here>"
}

• Response (JSON):

{
  "recipient_id": "<User id>",
  "text": "<Bot response>"
}

---

📄 Resume Scorer

This script scores a candidate's resume against a job description (JD) based on a predefined list of Computer Science skills using TF-IDF weighting.

Location: src/score_resumes.py

🚀 What It Does

• Uses NLP library spaCy to extract relevant skills from both JD and resume.
• Trains and loads a TF-IDF model based on job descriptions to weigh skill importance.
• Compares resume skills with JD-required skills.
• Uses a predetermined vocabulary list of CS related job skills.
• Calculates:
  • 🎯 A normalized match score (0 to 1).
  • ✅ A list of matched skills with weights (sorted by importance).
  • ❌ A list of missing but important skills (also sorted).

📦 Output Example

🎯 Score: 0.52
✅ Matched Skills (with weights): {'python': 0.26, 'sql': 0.34}
❌ Missing Skills (with weights): {'kubernetes': 0.90, 'r': 0.0}