DVC EC Pilot

Demo

Inspiration

Students often hear about extracurricular opportunities too late or struggle to determine whether an opportunity is legitimate, affordable, or relevant to their goals. Important information is scattered across websites, emails, and social media. We wanted to create one simple tool that helps students find opportunities and make better decisions.

What it does

PathFinder helps students discover competitions, volunteer opportunities, jobs, scholarships, and clubs. Students can filter opportunities by category, deadline, cost, and eligibility.

Students can also paste an activity link or description. The AI summarizes the opportunity, identifies important details such as the organizer, location, deadline, requirements, and fees, and provides ratings for authenticity, leadership potential, personal fit, and overall value.

How we built it

We built the MVP using Jac in a VS Code WSL environment. Jac models activities as structured objects with fields such as category, deadline, cost, and overall rating. We also created functions for displaying activities and filtering them by category.

The project is designed to use Jac and Jaseci AI capabilities for extracting information from pasted content, identifying missing or suspicious details, and generating personalized recommendations.

Challenges we ran into

One challenge was learning how to structure an application in Jac while keeping the data organized and easy to extend. We also had to think carefully about how to score authenticity without claiming that an AI system can guarantee that an opportunity is legitimate.

Another challenge was designing recommendations that are personalized rather than based only on popularity. Students have different interests, schedules, budgets, and long-term goals.

Accomplishments that we're proud of

We created a working Jac prototype that displays structured extracurricular activities and filters them by category. We also established the foundation for a personalized opportunity platform that combines discovery, summarization, and evaluation in one workflow.

What we learned

We learned how Jac can represent application data with typed objects and functions while providing a foundation for AI-native features. We also learned that useful recommendations need transparent explanations. Students should understand why an opportunity receives a certain rating instead of seeing only a final score.

What's next for PathFinder

Next, we plan to add student profiles, real activity search, saved opportunities, deadline reminders, and AI analysis of pasted links and documents. We also want to improve verification by checking official sources and clearly separating confirmed information from AI-generated estimates.