Skip to main content
Back to Projects
FULL-STACK PROJECT

LoksewaPro — AI-Powered Exam Platform

Full-Stack Developer & QA Tester

LoksewaPro — AI-Powered Exam Platform – project screenshot

Project Overview

Built a full-stack AI-driven platform that transforms uploaded question paper images into interactive practice quizzes. Features include intelligent OCR that reads both Nepali (Devanagari) and English text, regex-powered MCQ parsing that extracts questions with options A–D, automatic answer key matching, and real-time quiz scoring with detailed performance analytics.

Problem Statement

Loksewa aspirants in Nepal lacked a digital platform to practice with real past question papers, especially those in Nepali language, requiring an OCR-based solution.

Key Features

  • OCR-based question paper scanning
  • Bilingual OCR (Nepali Devanagari & English)
  • Automatic answer key matching
  • Real-time quiz scoring with analytics
  • Drag-and-drop upload interface
  • Camera capture support for mobile
  • Interactive quiz engine with timer
  • Skip/return and review mode

QA & Testing Details

Tested OCR accuracy across diverse paper qualities, lighting conditions, and font variations. Validated regex parsing for 20+ MCQ formatting edge cases. Performed cross-browser testing for consistent OCR canvas rendering. Tested camera capture flow on mobile devices. Verified quiz scoring logic and answer key matching.

Screenshots

LoksewaPro — AI-Powered Exam Platform screenshot 1