ResumeModifier

A Flask-based backend service for fetching, storing, and matching job postings with resume descriptions using semantic search powered by OpenAI.

---

Features

• Fetch jobs from RapidAPI (JSearch API)
• Store jobs in MongoDB (with OpenAI vector embedding)
• Match jobs to user resumes using cosine similarity
• Preloading strategy for fallback when live API data is insufficient
• Dockerized for easy setup

---

Requirements

• Docker
• Docker Compose
• OpenAI API Key (for embedding)
• RapidAPI Key (for job data)

---

Setup & Usage

1. Clone this repo

git clone https://github.com/TaoJR/ResumeModifier.git
cd ResumeModifier

2. Create your .env file

In the root directory:

echo "OPENAI_API_KEY=sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" > .env

Replace the value with your actual OpenAI key.

Do NOT commit .env to GitHub.

3. Build and start the containers

docker-compose up --build

This will:

• Start the Flask app on localhost:5000
• Start MongoDB (available at mongodb://localhost:27017 for Compass, or mongo:27017 inside Docker)

---

API Endpoints

/jobs/fetch_jobs (GET)

Fetch jobs from API and store them with embedding.

Params (as query):

• query: Job keyword (e.g., Python)
• max_pages: Number of pages (default 1)

curl "http://localhost:5000/jobs/fetch_jobs?query=Python&max_pages=1"

---

Match Jobs API

Matches the user’s skills with top-k job listings based on OpenAI embedding similarity.

Endpoint: POST /api/match_jobs
Content-Type: application/json

Request Body:

{
  "candidate_keywords": "Python Flask MongoDB backend",
  "top_k": 5
}

curl -X POST http://localhost:5000/api/match_jobs ^
  -H "Content-Type: application/json" ^
  -d "{\"candidate_keywords\": \"Python Flask MongoDB backend\", \"top_k\": 5}"

---

Architecture

• app/job_api.py: Job fetching & embedding logic
• app/routes.py: Resume-job matching logic
• app/database.py: MongoDB connection
• matcher.py: Core cosine similarity logic
• run.py: Entrypoint

---

Testing MongoDB (optional)

You can connect with MongoDB Compass:

• Host: localhost
• Port: 27017
• Database: jobSearch
• Collection: jobs

---

Contributors

• @TaoJR — Backend / Docker / Matching logic

---