Personalized Matching

A three-stage matching pipeline

Stage 1

Career-track classification

Your profile is classified into one of six broad tracks — Technical, Medical, Creative, Business, Research, or Vocational — used to narrow the search space before scoring.

Stage 2

Collaborative filtering

Students with a similar profile to yours inform which opportunities tend to be worth surfacing — a signal that gets stronger as more students use the platform.

Stage 3

Weighted scoring

Each opportunity is scored against your academic performance, interests, skills, career-track alignment, location, and eligibility, weighted differently depending on whether it's a scholarship, college, internship, or job.

How the final score is built

The three stages combine into one composite score you can see for every recommendation:

Composite Score = 0.5 × Weighted Score + 0.3 × Collaborative Score + 0.2 × Recency Score

Weighted scoring carries the most weight because it works well from day one, using nothing but your own profile. Collaborative filtering adds more value as the platform's user base grows.

Where we're honest about the current state: the career-track classifier is currently a transparent, rule-based system — not yet a model trained on real outcome data, since that data doesn't exist yet for a platform this new. It's built to be swapped for a trained model as real usage accumulates, without changing anything else in the pipeline.

You can see why you were matched

Every recommendation shows its score breakdown — how much came from your profile fit versus peer signal versus deadline urgency — so matching never feels like a black box.