JobMate

Auto Apply 2.0: Powered by JobMate

JobMate is a two-sided recruitment platform designed to streamline hiring for recruiters while supporting applicants with ATS-friendly resume optimization. Built on the foundational work of the AutoApply app, which focused on AI-based job recommendations and resume tailoring, JobMate extends functionality to include full recruiter-side tools and end-to-end applicant management.

How the Previous App Worked: AutoApply

AutoApply is a resume/job matching and analysis platform powered by AI. It processes resumes and job descriptions, evaluates compatibility, and manages data using MongoDB, Streamlit, and Apache Airflow.

Features

• Upload and analyze resumes and job descriptions (PDF)
• Extract and evaluate skills, experience, and compatibility
• Store and manage resumes in MongoDB
• Orchestrate data pipelines with Apache Airflow
• Modern UI with Streamlit

---

New Features in JobMate

Applicant Features

• ATS Compatibility Scoring – Evaluates resume alignment with job descriptions and provides actionable recommendations.
• Resume Optimization – AI-powered tailoring to improve match scores based on job requirements.
• Cover Letter Generation – Creates personalized, ATS-friendly cover letters matched to tailored resumes.
• Ghost Job Detection – Identifies fake or inactive job postings to protect applicants.
• AI-Powered Interview Questions Generator – Generates interview questions based on skills and job descriptions.

Recruiter Features

• Job Posting Management – Create, update, pause, and remove job postings.
• Candidate Management System – Update applicant status in the recruitment pipeline (Talent Journey).
• AI Salary & Skills Recommendation – Suggests competitive salaries and in-demand skills.
• AI-Powered Interview Question Generation – Generates role-specific interview questions with scoring rubrics.

Job Auto Application Feature

• L.i.n.k.e.dIn E@sy @pply Automat0n – Autom@tes the E.a.s.y_@pply process, reducing manual form-filling.
• G.l.a.s.s.door Q.u.i.c.k @pply Automat0n – Autom@tes Glass_door applications with smart job filtering.
• Multi-Step Form Handling – Navigates and completes multi-page application flows with file uploads and questions.
• Application Tracking – Saves applied job details (title, date, URL) to CSV.

Prerequisites

• Docker and Docker Compose
• (Optional) Python 3.9+ for local development
• Get Google Gemini API Key (https://ai.google.dev/gemini-api/docs/api-key)

---

1. Clone the Repository

git clone <your-repo-url>
cd JobMate

---

2. Create Environment Variables

Create a .env file in the project root (not inside src/). Example:

# MongoDB
MONGODB_URI=mongodb://airflow:airflow@mongodb:27017/autoapply?authSource=admin
MONGODB_DB=autoapply
MONGODB_COLLECTION=resumes

# Gemini API Key
GEMINI_API_KEY=your_gemini_api_key_here

# Airflow/Postgres (defaults)
POSTGRES_USER=airflow
POSTGRES_PASSWORD=airflow
POSTGRES_DB=airflow

# SonarQube
SONAR_TOKEN=your_sonarqube_token_here
SONAR_PROJECT_KEY=JobMate

# Streamlit
ENVIRONMENT=development
DEBUG=True
LOG_LEVEL=INFO

Replace your_gemini_api_key_here with your actual Gemini API key.

---

3. Build and Start the Project

docker-compose up -d --build

This will start:

• MongoDB (database)
• Postgres (for Airflow)
• Apache Airflow (webserver and scheduler)
• Streamlit app (web interface)

---

4. Access the Applications

• Streamlit App: http://localhost:8501
• Apache Airflow: http://localhost:8080
  • Username: airflow
  • Password: airflow
• MongoDB: localhost:27017 (use a MongoDB client)

---

5. Upload and Analyze Resumes

• Go to the Streamlit app
• Upload your resume and job description (PDF)
• View compatibility analysis and manage resumes in the database

---

6. Stopping the Project

docker-compose down

To remove all data (including database volumes):

docker-compose down -v

---

7. Troubleshooting

• Ensure your .env file is in the project root
• Check container logs with docker-compose logs <service>
• If you change environment variables, restart the containers

8. For L.i.n.k.e.dIn Aut0_Apply

• This automation runs as a standalone Python script and should be executed in a Python-compatible IDE (e.g., VS Code, PyCharm).
• pip install -r requirements.txt
• Have have a fully completed L1nked1n profile with CV and Phone number attached
• Ensure you have a chrome browser, and updated
• To activate the script run: python app.py
• Enter credentials, job title and location
• For every succesful application, a csv for application tracking is automatically saved on the same folder

---

flowchart TD
    A[Start Script] --> B[User Inputs LinkedIn Credentials, Job Title, Location]
    B --> C[Log in to LinkedIn]
    C --> D[Go to Jobs Page & Apply Easy Apply Filter]
    D --> E[Load All Job Cards on Page]
    E --> F{Already Applied?}
    F -- Yes --> G[Skip Job]
    F -- No --> H{Easy Apply Available?}
    H -- No --> G
    H -- Yes --> I[Start Application Modal]
    I --> J[Click Next / Review / Submit]
    J --> K{Page Progresses?}
    K -- No --> L[Discard Application & Close Modal]
    K -- Yes --> M[Submit Application]
    M --> N[Log Success to CSV]
    N --> O[Move to Next Job]
    O --> E

9. For Glass_d0or Aut0_apply

• This automation runs as a standalone Python script and should be executed in a Python-compatible IDE (e.g., VS Code, PyCharm).
• pip install -r requirements.txt
• Have a fully completed Glass_d0or profile with CV and Phone number attached
• Most updated chrome browser is also required
• Set up a proxy account with Bright Data -Obtain host and port details, install ssl certificate and activate
  • Put in the proxy details in setup_proxy.py
• Open the config.json and update with your credentials and search preferences
• To activate the script run: python runner.py
• For every succesful application, a csv for application tracking is automatically saved on the same folder

flowchart TD
    A1[Start Script] --> B1[User Inputs Glassdoor Credentials, Job Title, Location]
    B1 --> C1[Log in to Glassdoor]
    C1 --> D1[Enable Easy Apply Filter]
    D1 --> E1[Load Job Cards]
    E1 --> F1{Already Applied?}
    F1 -- Yes --> G1[Skip Job]
    F1 -- No --> H1[Open Job Application Page]
    H1 --> I1[Fill Out Application Form]
    I1 --> J1[Upload Resume & Documents]
    J1 --> K1[Answer Screening Questions]
    K1 --> L1[Review Application]
    L1 --> M1[Submit Application]
    M1 --> N1[Log Success to CSV]
    N1 --> O1[Move to Next Job]
    O1 --> E1

---

10. Project Structure

JobMate/
├── src/                  # Source code (Streamlit, utils, etc.)
├── resume/               # Resume data (mounted in container)
├── data/                 # Data files
├── logs/                 # Airflow logs
├── plugins/              # Airflow plugins
├── Dockerfile.streamlit  # Streamlit Dockerfile
├── docker-compose.yml    # Docker Compose config
├── sonar-project.properties # SonarQube configuration
├── run-sonar.sh         # SonarQube scanner script
├── .env                  # Environment variables
└── README.md             # This file

11. Environment Variables Reference

Variable	Description
MONGODB_URI	MongoDB connection string
MONGODB_DB	MongoDB database name
MONGODB_COLLECTION	MongoDB collection name
GEMINI_API_KEY	Gemini API key
POSTGRES_USER	Airflow/Postgres user
POSTGRES_PASSWORD	Airflow/Postgres password
POSTGRES_DB	Airflow/Postgres database
SONAR_TOKEN	SonarQube authentication token
SONAR_PROJECT_KEY	SonarQube project key identifier
ENVIRONMENT	App environment (development/prod)
DEBUG	Debug mode (True/False)
LOG_LEVEL	Logging level (INFO/DEBUG/WARN)

---

12. Code Quality with SonarQube

This project includes SonarQube integration for code quality analysis. SonarQube helps identify bugs, code smells, and security vulnerabilities in your code.

To run a SonarQube analysis:

docker-compose run --rm sonar-scanner

The analysis results will be available on your SonarQube server at: https://sonarqube.nunchisolucoes.com/dashboard?id=JobMate

For configuration details, see sonar-project.properties file. The scanner uses environment variables from your .env file.

11. Useful Commands

• View logs: docker-compose logs <service>
• Rebuild containers: docker-compose up -d --build
• Stop all: docker-compose down
• Remove all data: docker-compose down -v
• Run SonarQube analysis: docker-compose run --rm sonar-scanner

---

License

This project is solely for personal or educational purposes only

Authors

• Jennylynne Dominguez
• Ricardo Tassio Dantas Da Silva
• Bo Yang
• Chaw Su Su Thin
• Joyce Ann Murillo
• Mugilmithran Kathiravan
• Johan Rodriguez Rodriguez
• Andres Felipe Santa Lozano
• Hazel Portia Elaine Santos

Acknowledgements

This project was developed in faithful completion of:
Big Data Capstone Project
Lambton College, Mississauga
Course: 2025S-T3_BDM 3035_01
Professor: Bhavik Gandhi