Team 102_CareerLaunch

Inspiration

We noticed how fragmented the job search process is. Opportunities are spread across multiple platforms such as Indeed, Glassdoor, JobStreet, and others. Keeping track of saved jobs, applications, and required skills quickly becomes overwhelming.

Our solution is to build a single platform that brings all job listings together, helps users stay organised, and turns job searching into a more structured and less stressful experience.

What it does

Career Launch is a unified job discovery and application tracking platform.

It allows users to:

Browse job listings aggregated from multiple platforms
Search and filter jobs by keyword, source, type, and location
Track application statuses (Not Applied, Applied, Interview Scheduled, Offer Received, Rejected)
Generate a personalized upskilling roadmap based on skills extracted from saved job descriptions

The roadmap helps users understand what skills are most in-demand across the jobs they are interested in and provides guidance on what to learn next.

How we built it

The backend was built using Python and Flask, exposing REST-style API endpoints for jobs, saved listings, applications, and roadmap generation.

Job data is loaded from multiple CSV datasets and normalized into a unified structure using Pandas. Each job is assigned a unique identifier (job ID) to ensure consistent tracking across saving and application features.

We used SQLite as a database to persist:

Saved jobs
Application status
Timestamps for user actions (such as save/update)

For the roadmap feature, job descriptions are analyzed using regular expressions to detect common technical skill keywords. These skills are aggregated to generate a personalized learning roadmap.

The frontend communicates with the backend through the browser Fetch API. The UI focuses on clarity, smooth navigation, and visual feedback so users can easily manage large volumes of job listings.

Challenges we ran into

One challenge we had to normalize data from multiple job sources with inconsistent column formats, which required careful handling to prevent missing or broken listings.

What we learned

Through this project, we learned how to:

Design REST APIs that support real user workflows
Normalize and clean real-world datasets
Maintain application state between frontend and backend
Translate a real problem into a usable product under time constraints

Most importantly, we learned how thoughtful product design can significantly improve the user experience in seemingly mundane and scary tasks like job searching.