cd .. / projects
<scholaris/>

Scholaris

IN PROGRESS
Scholarship-Matching Platform•2026•Full-Stack Developer•Personal Project
Scholaris primary preview
<about-project/>

Scholaris (in progress — currently an early Flutter/Dart implementation foundation, not yet released) tackles a real problem: students usually scan dozens of scholarship postings by hand with no way to tell which ones they are actually eligible for. What exists today is the project foundation: a Flutter/Dart codebase with the initial application structure in active development. The planned direction is to model each opportunity with structured eligibility rules and compare them against a student's profile — course, year level, and background — using a transparent scoring layer so the most relevant matches surface first, with Supabase (PostgreSQL) handling authentication and data. Those capabilities are designed and planned, but not yet implemented.

<tech-stack/>
<key-features/>
Flutter/Dart project foundation — in active development (current)
Initial application structure — in active development (current)
Supabase/PostgreSQL integration (planned)
Authentication (planned)
Scholarship data model and catalog (planned)
Eligibility/profile matching and match scoring (planned)
Bookmark and save opportunities for later (planned)
Administrator publishing workflow for new scholarships (planned)
Android, iOS, and web from a single Flutter codebase (planned)
<technical-decisions/>

Why Flutter with a single Dart codebase?

Students search for scholarships mostly from phones, so the plan is to ship Android, iOS, and web from one Dart codebase to maximize reach while keeping the UI consistent. The Flutter/Dart foundation is what is being built now.

Why Supabase with Row-Level Security? (planned)

The planned backend pairs Supabase (PostgreSQL) with Row-Level Security to segment students and administrators at the database layer, and real-time subscriptions would keep the opportunity list fresh without custom push infrastructure. This is design intent for the upcoming backend work, not yet implemented.

<what-i-learned/>
01

Structured eligibility beats free-text matching

Planned design principle for the matching engine: typed criteria (course, year level, location, financial status) should make ranking deterministic and explainable, where fuzzy text comparison is neither. This shapes the upcoming matching work.

02

Explicit scoring is easier to tune

Planned design principle for the scoring layer: a transparent weighted score should make it straightforward to adjust which criteria matter most without rewriting the matching engine.