Skip to content
RUCHITHA- home

Command menu

Jump to a page, open a link, or change a setting.

Projects

AcadPulse

HomeIn progress

Academic analytics, predicted

AcadPulse screenshot
  • Next.js
  • React
  • TypeScript
  • Chart.js
  • D3.js
  • Anthropic API

AcadPulse is an academic analytics platform built around a genuinely hard prediction problem: forecasting a student's CGPA from incomplete, messy academic history - scanned marksheets, inconsistent formats, partial semesters.

Predictions come from an ensemble (Random Forest + Polynomial regression + EWMA) tuned to 92% confidence, fed by an AI pipeline that extracts grades directly from PDF or image marksheets rather than requiring manual entry.

The output side is built for actually understanding the data, not just displaying it: 7 chart types, student-to-student comparison, and one-click PDF report generation.

Highlights

  • ML ensemble (Random Forest + Polynomial + EWMA) at 92% prediction confidence
  • AI-powered PDF/image marksheet extraction - no manual data entry
  • 7 chart types, student comparison and PDF report generation