CareerCompass

CareerCompass homepage, where users customize priorities
CareerCompass result calculator score
Regret probability bar chart average by major
Scatter plot comparing starting salaries against job satisfaction scores

Inspiration

Choosing a college major is one of the largest financial decisions most people make when they are 18 years old, yet it’s often based on passion, pressure, or incomplete information.

The reality:

The average student graduates with ~$37,000 in debt
Nearly 1 in 3 graduates regret their major within 5 years
AI and automation are reshaping entire industries

We wanted to build a tool that combines:

Major choice
Financial risk
Labor market data
AI disruption
College choice impact

CareerCompass was created to answer one powerful question:

“Will I regret this major at this school?”

What It Does

CareerCompass calculates a Regret Probability Score (RPS) for 25+ college majors using real economic and institutional data.

The RPS factors in:

Starting Salary
Average Student Debt
10-Year Job Growth
Automation Risk
Unemployment Rate
Job Satisfaction
Graduate Degree Requirement
College Choice
Debt-to-Income Stress Penalty

Users can:

Customize how much each factor matters to them
Select their university
Visualize salary vs happiness
Analyze automation risk vs job growth
Receive a personalized Freshman Action Plan

CareerCompass reframes choosing a major as a strategic decision, not just an emotional one.

How We Built It

Frontend & Framework

Built using Streamlit
Custom CSS design system for a clean, modern UI
Multi-tab layout for calculator, comparison, insights, and data transparency

Data & Sources

National Association of Colleges and Employers (NACE) – 2024 Salary Survey
U.S. Bureau of Labor Statistics (BLS)
U.S. News & World Report - Best National Universities Rankings 2026

We integrated university ranking data to apply a tiered prestige modifier, allowing the same major to yield different regret scores depending on institutional strength.

Scoring Engine

We designed a normalized scoring system that:

Scales each metric between 0–1 using reference ranges
Applies user-defined weights (must sum to 1.0)
Calculates a weighted regret baseline
Outputs a final 0–100 Regret Probability Score

All recalculations occur in real time when users adjust weights or switch schools.

Challenges We Ran Into

Quantifying “College Choice” Without Overpowering the Model

We had to ensure university prestige influenced outcomes realistically without dominating the entire score. We built tier-based modifiers instead of linear scaling.

Normalizing Mixed Data Types

Economic, satisfaction, and automation metrics all operate on different scales. We carefully designed normalization logic to preserve fairness.

Weight Customization Edge Cases

Allowing users to fully control weights introduced extreme configurations that distorted scores. We implemented balancing checks and guardrails.

Designing a Regret Framework

Regret is psychological — but we had to model it mathematically using economic risk signals.

Accomplishments We're Proud Of

Built a fully custom regret probability algorithm
Successfully integrated college choice impact into the scoring model
Enabled real-time personalization through adjustable weights
Designed clear economic visualizations for complex data
Created actionable freshman-level guidance

Most importantly:

We transformed an abstract life decision into a structured, data-driven risk model.

What We Learned

Major choice alone doesn’t tell the whole story, school choice materially changes outcomes.
Students undervalue debt-to-income risk.
Personalization increases engagement dramatically.
Data transparency builds trust.
Visual storytelling is critical when communicating financial risk.

We also learned that regret isn’t binary — it’s probabilistic, contextual, and deeply tied to both what you study and where you study it.