Team59_CareerGPS

In the sea of job hunting, we're your compass

Comment

Inspiration

Our inspiration came from the fragmented nature of the SG job market. We noticed that many fresh graduates and professionals felt overwhelmed by the sheer volume of job postings across different platforms. We wanted to bridge this gap by unifying these listings into a single source. Moreover, we were deeply moved by the stories of our seniors at NTU who created their startup to help us prep for technical interviews. We wanted to "automate" this idea into an AI-driven roadmap that makes upskilling feel like a clear path.

What it does

CareerGPS is a one-stop career navigation hub tailored for the Singaporean workforce.

Smart Aggregation: It scrapes and centralizes job postings from major local boards, removing the need for cross-platform hopping and reduces time looking for desired roles.

Tailored Filtering: Instead of generic keyword searches, it uses AI to match roles to your specific profile and aspirations.

Dynamic Roadmaps: It identifies the "skill gap" between your current resume and your dream job, generating a step-by-step learning path to help you upskill efficiently.

How we built it

We treated this like a full stack project, beginning with the UI from Figma. Then, we cleaned and adjusted the datasets into one large dataset accommodating everyone.

After this, we integrated our own AI agent that would interact with users 24/7 for career advice.

Finally we integrated all of this together and came up with a majorly bootstrapped startup idea.

Challenges we ran into

Being Year 1's in university, we found a lot of tools and languages new and unfamiliar cause a bit of a challenge. However, we soon overcame it with a positive attitude and a hunger to learn.

Also In Singapore, job titles can vary wildly (e.g., "AI Engineer" vs. "Machine Learning Specialist"). Normalizing these to map them correctly to skills required significant fine-tuning of our NLP model.

We also struggled with getting the Resume Parsing code to work, though we did figure it out a while later.

Accomplishments that we're proud of

Precision Matching: Achieving a high "Relevance Score" when testing our tool against actual local job postings for Data Science and AI roles.

Localized Context: Integrating specific Singaporean industry requirements, such as focusing on certifications that are highly valued in the local tech ecosystem.

What we learned

Data Quality > Model Complexity - cleaning and structuring our scraped job data was far more important than the AI model

Through feedback, we realized that job seekers don't just want more jobs; they want clarity on why they aren't qualified yet and how to get there.

What's next for 59_CareerGPS

We're planning to improve our job matching algorithm to include more variables so we can venture into the global job market too.

We aim to map our roadmaps directly to SkillsFuture-subsidized courses, making it easier for users to find affordable training.

Mock Interview Simulation: Building a module where users can practice technical interview questions specifically for the jobs they are tracking.

Share this project:

Updates