Career Pal

Career Pal homepage
Input question
Multiselect question
Career options
Career option in-depth analysis
Sign up prompt
Tasks list with categories filtering feature
Task details, including description and helpful resources

Inspiration

As students, many of us feel lost and overwhelmed when it comes to choosing a career. We’re told to figure it out, but rarely shown where to start. The pressure to succeed, paired with a competitive job market, makes it even harder.

As an NTU undergraduate, I’ve felt this firsthand — unsure of what roles suit me or who to ask for help. That’s when the idea struck: what if we had an AI-powered career mentor that understands our personality, goals, and gives us a clear plan forward?

That’s the heart of Career Pal — helping students navigate their career journey with clarity, confidence, and support.

What it does

Career Pal guides users through five simple steps:

Step 1: Career Discovery Test – Users complete a short test by answering a series of questions. The responses are then analyzed.
Step 2: Career Recommendations – Based on the analysis, the top 3 most suitable careers are recommended, tailored to each user's values, personality, and interests.
Step 3: Career Exploration – Users explore each suggested role in detail, including job responsibilities, required skills, salary range, growth outlook, and potential trade-offs.
Step 4: Account Creation – After choosing a preferred career, users create an account to receive a personalized career roadmap.
Step 5: Unlock Additional Features – As users complete tasks, they unlock access to advanced features such as resume building, job search tools, mentorship, and career coaching. (Note: These features are not included in the current MVP.)

How we built it

Career Pal is powered by AI through the DeepSeek model via OpenRouter API.
The frontend is built with React and Redux Toolkit (RTK) for scalable state management in the future. We use Express.js to handle backend logic and communication between the frontend and AI model. We used Firebase Auth for secure user authentication.

Challenges we ran into

With only two of us on the team, focus and commitment were crucial.
Unexpectedly, one teammate fell sick with the flu mid-hackathon.

But thanks to a well-structured plan and determination, we powered through and delivered a product we’re proud of.

Accomplishments that we're proud of

Seamless AI model integration using OpenRouter
Scalable Redux state management
Functional Firebase authentication setup

What we learned

How to implement Firebase Authentication
Crafting clear and structured prompt engineering to instruct the AI model to perform analysis
Managing frontend-backend communication in a full-stack app