AI Resume Screening System

AI Resume Screening System is a portfolio-ready Flask application that parses resumes, extracts skills and keywords with NLP, compares them to a job description, and ranks candidates using a hybrid scoring approach.

Features

• Upload one job description and multiple resumes
• Extract skills from resumes and job posts using spaCy phrase matching
• Generate keywords from cleaned text with TF-IDF
• Score candidates using a hybrid of skill overlap and text similarity
• Store jobs, resumes, and ranking results in SQL via SQLAlchemy
• Use PostgreSQL in production or SQLite for quick local development
• Access both server-rendered pages and JSON API endpoints

Tech Stack

• Python
• Flask
• spaCy
• Scikit-learn
• SQLAlchemy
• PostgreSQL

Project Structure

app/
  services/
  static/
  templates/
data/
samples/
tests/
run.py

Local Setup

1. Create a virtual environment and activate it.
2. Install dependencies:

   pip install -r requirements.txt

3. Copy .env.example to .env and update values if needed.
4. Optional but recommended: install the spaCy English model:

   python -m spacy download en_core_web_sm

   The app still runs with a lightweight fallback tokenizer if the model is not installed.
5. Start the app:

   python run.py

The app will create the SQLite database automatically if DATABASE_URL is not set.

PostgreSQL Example

DATABASE_URL=postgresql://postgres:postgres@localhost:5432/resume_screening
SECRET_KEY=replace-this
UPLOAD_FOLDER=uploads
SPACY_MODEL=en_core_web_sm

SQLite Fallback Example

SECRET_KEY=local-dev-secret
DATABASE_URL=sqlite:///resume_screening.db
UPLOAD_FOLDER=uploads
SPACY_MODEL=en_core_web_sm

Sample Workflow

1. Create a job description from the dashboard or POST /api/jobs.
2. Upload multiple .txt or .pdf resumes from the dashboard or POST /api/resumes/upload.
3. Run the screening flow from the dashboard or POST /api/screenings/run.
4. Review ranked results on the UI or via GET /api/screenings/<job_id>/results.

API Examples

Create a job:

curl -X POST http://127.0.0.1:5000/api/jobs \
  -H "Content-Type: application/json" \
  -d "{\"title\": \"Python NLP Engineer\", \"description\": \"Need Python, Flask, PostgreSQL, spaCy, and scikit-learn.\"}"

Upload resumes:

curl -X POST http://127.0.0.1:5000/api/resumes/upload \
  -F "resumes=@samples/resumes/alex_johnson.txt" \
  -F "resumes=@samples/resumes/sam_lee.txt"

Run screening:

curl -X POST http://127.0.0.1:5000/api/screenings/run \
  -H "Content-Type: application/json" \
  -d "{\"job_id\": 1}"

Get results:

curl http://127.0.0.1:5000/api/screenings/1/results

Screenshots

• Add dashboard screenshot here
• Add ranking results screenshot here

Sample Data

• samples/job_description.txt
• samples/resumes/alex_johnson.txt
• samples/resumes/sam_lee.txt

Testing

Run:

pytest