AdmitGPT is a transparent, mathematically rigorous, and fully client-side college admissions analytical engine. It rejects "black box" algorithms in favor of a Gated Multiplicative Model that balances academic hard-gates with high-impact "Spike" achievements.
- Gated Multiplicative Architecture: Enforces academic rigor while allowing outliers to shine through specialized "Spike" modeling.
- Identity & Verification Engine: Every result is cryptographically signed. Standard students receive an "Audit Signature," while high-achievement outliers receive a premium "Candidate Excellence Certificate."
- Full Privacy Protection: All calculations run locally in your browser. Sensitive data can be masked while maintaining a valid cryptographic audit trail.
- Transparency First: Every formula, weight, and limitation is disclosed. Access the Transparency Engine to see the math behind the numbers.
- Peer Invitation System: High-achievement outliers gain a direct "Peer Invitation" channel to the creator for personal mentorship and story-sharing.
- Core: Next.js (React), TypeScript, TailwindCSS
- Security: AES-256 CTR (CryptoJS) for Audit Signatures
- PDF Core: jsPDF for secure, cryptographically-signed report generation
- Math Engine: Multi-dimensional probability modeling with k-NN clustering across a 5.5MB dataset (~10,000+ profiles)
-
Install Dependencies:
npm install
-
Run Development Server:
npm run dev
-
Production Build:
npm run build
We ensure mathematical integrity through a rigorous testing suite:
npm run testSee tests/comprehensive.test.ts for logic verification of Major Competitiveness, International Anomalies, and Hard Gates.
Built by student, for students. Zero Secrets. No Black Boxes.