2025 · Sole developer
Tuition Forecasting & Financial Planning
A predictive model that forecasts multi-year university costs and turns them into a personalized savings plan, validated to ±5% accuracy against a decade of real data.
The problem
Students and families, myself included, constantly underestimate what an engineering degree actually costs over four-plus years, tuition doesn’t stand still, and neither does everything else attached to it. I wanted a tool that takes messy historical tuition data and turns it into an honest forecast, and then into an actual plan.
What I built
- Requirements first. Before writing a line of modeling code, I ran a proper requirements analysis to nail down exactly what the forecast needed to answer. Skipping this step is how you end up with a beautiful model that answers the wrong question.
- Data validation at scale. I pulled and analyzed 10+ years of historical university cost data to calibrate the model and test how well it stood against actual reported tuition rates.
- A model you can actually trust. It lands within ±5% variance of real outcomes against actual reported tuitions(courtesy of some of my friends + the government of Canada’s reported data)close enough to plan around, which was the whole point.
- From numbers to decisions. Forecasts are only useful if they turn into action, so I built spending-pattern algorithms that convert the projection into a personalized savings plan, visualized through clear Matplotlib dashboards.
What it demonstrates
I owned a data problem end to end: framing the question, sourcing and validating the data, building and calibrating the model, and reporting a result someone could actually make a decision from. More than a notebook that just runs. Something that answers a question people genuinely have.