How We Built an AI Career Path Recommender

It started with a Google Form. One minute we were zoning out in class, the next we were signing up for the ICTAK Techathlon 2024-25. The theme? NextGenAI. The location? Infopark, Koratty. The vibe? Pure chaos with a side of ambition.

About That Team Name...

Let me explain "Team Rizzler." It wasn't planned. At all. When we were filling out the registration form in between classes, our sleep-deprived brain thought it would be hilarious to put "Rizzler" as a placeholder until we came up with something better. Except we forgot to change it. Completely forgot.

Flashback to the initial roll call:

"And next up... Team Rizzler?"

The organizers paused, double-checked the list. We sheepishly raised our hands as the room burst into laughter. A meme was born before the hackathon even started. No regrets.

The Problem: Course Selection Feels Like Gambling

You know that feeling when you're staring at course options wondering which one won't destroy your future? Yeah, that's what we wanted to fix.

We built an AI-powered FYUGP Course Pathway Recommender. Essentially:

"We read your mind so your academic counselor doesn't have to."

The Build

Here's what went into our solution:

• Psychometric testing (because your mental health matters)
• Holland career codes (those personality tests you secretly enjoy)
• Academic record analysis (because past performance does matter)
• AI/ML prediction models (the secret sauce)

All wrapped in a UI so clean your grandmother could use it.

What It Actually Does

Our system doesn't just throw random courses at you. It:

1. Predicts the best FYUGP track based on your unique profile
2. Identifies skill gaps you might have
3. Suggests additional certifications to boost your prospects
4. Visualizes your potential career paths

"It's like if Spotify and a career counselor had a baby that went to college."

The Tech Stack

We built this with:

• Python backend (Google Colab for training and fine-tuning our AI models)
• NLP for processing what students actually mean
• Basic frontend (React, because we like to suffer)
• Azure for deployment (using student credits)

The Team Members :

• Me (Mishal Shanavas)
• Mathew Geejo
• Shayen Thomas

The Victory Moment

When they called out "2nd Runner-Up" and then struggled to pronounce "Rizzler" with a straight face, we experienced that rare moment when imposter syndrome takes a break.

The judges later told us they loved our project but were equally entertained by our team name presentation. Apparently, our awkward explanation of "it was an accident" added charm points.

For the official announcement, check out this LinkedIn post.

This post is part of my Projects series – documenting the weird and wonderful things I build instead of sleeping (and accidentally naming).