JobTrack

JobTrack - Stop searching. Start matching!

Comment

JobTrack - Stop searching. Start matching!

Inspiration

Job hunting is overwhelming. Candidates apply to roles without knowing if they truly fit, while recruiters sift through mountains of resumes.
JobTrack answers a simple question:
What if matching talent to opportunity could be intelligent, automated, and instant?

It analyzes resumes, understands job postings, and connects the dots using contextual matching.

What It Does

JobTrack automatically:

  • Analyzes resumes and parses job postings
  • Compares skills using a contextual matching model
  • Computes match scores between candidates and jobs
  • Outputs clean, rankable results for dashboards or applications

It focuses on compatibilities, making matches more accurate and fair.

How We Built It

  1. Job Parsing – Normalizes and structures infromation about job postings, stores position information into relational database.
  2. Resume Analysis - Receives resume and analyzes with developed ML model, leverages skillset and experience extraction.
  3. Position-Resume Skill Matching - Compares position's and resume's skill descriptions using own ML model via contextual matching.
  4. API Routing - Ensures smooth operations of client with database, backend with ML, and data extraction and user authentication.
  5. User Interface - Implements client communication with API as well as modern and pleasant UI.
  6. AI Assistant - Provides personalized tips using user's resume context, which allows user to improve quality of the resume.
  7. Docker Conterization - Ensures portability of API services on any host machine.
  8. Database - PostgreSQL database, which stores user's authentication credentials, parsed jobs, resume records, and parsed job matchings.

Challenges

  • Messy real-world data and inconsistent skill formats
  • Complex database schemas and field mapping
  • Integrating ML models with structured inputs
  • Implementing communication of API and ML model services
  • Ensure consistent database tables and information storage

These challenges refined the architecture and strengthened the system.

Accomplishments

  • Designed and implemented a clean, modern web interface that makes job discovery intuitive and enjoyable
  • Built a seamless end‑to‑end flow between the frontend, API, database, and ML matching engine
  • Delivered a responsive, user‑friendly experience that feels polished, fast, and ready for real users

Potential Next Steps

  • Build recruiter and candidate dashboards
  • Use embeddings for deeper semantic matching
  • Deploy as a real-time API for job platforms

JobTrack is ready to evolve into a full intelligent job hunt assistant.

Technical Stack

  • Frontend: React, TypeScript, SCSS, React-Hot-Toast, Axios, Lucid-React-Icons
  • Backend: FastAPI, Python, SQLAlchemy, Docker, JWT, requests
  • ML: Python, Scikit-learn, SentenceTransformers, json, Llama
  • Database: PostgreSQL, Supabase
  • Parser: Playwright

Built With

Share this project:

Updates