Deep Hayer

projects

admittr

2026 — building

Data-grounded college admissions. Swipe to discover, see admission probability against real applicants.

stack: Next.js 14, TypeScript, Neon Postgres, Prisma, Tailwind, shadcn/ui

A “chance me” that does not rely on Reddit. Students fill a profile, get per-college admission probability estimates blending a strength score with similarity-weighted historical outcomes, and see up to 20 comparable applicants with their actual results.

Latest

Growing the applicant dataset and improving probability calibration. Transfer-student onboarding now handles GPA, credits, and transfer reasons separately.

Status

End-to-end flow working in dev. Calibration is the bottleneck; outcomes data is the input I am farming.