تفاصيل العمل

Designed and built Centery, a full-stack, multi-tenant educational platform that deeply integrates AI into core academic workflows. The platform serves administrators, instructors, assistants, students, and parents with role-based dashboards for academic management, exams, attendance, grading, reporting, messaging, and payments. Unlike traditional LMS solutions, Centery was engineered from the ground up to leverage AI across almost every high-value educational task, making learning, teaching, and assessment more efficient, intelligent, and data-driven.

The platform includes an AI-powered educational chatbot that provides session-aware support to students, answering questions using lesson video metadata, PDFs, text files, and course materials as context. Chat history is persisted in PostgreSQL, and usage limits are controlled by subscription plans. The platform also features automated rubric generation and AI-assisted written-answer grading, capable of evaluating both typed and image-based submissions using Groq vision models. AI-generated exam questions are supported for MCQ and written formats, powered by Groq LLMs.

For exam security, Centery implements AI-based proctoring using YOLOv8 and OpenCV to detect people and phones from camera frames in real time. The system tracks browser and tab events, captures screenshots, logs suspicious activity, and provides alert review dashboards for instructors. The platform also supports OMR-style bubble sheet printing and batch grading for MCQ exams. All AI services are implemented as separate Python microservices (FastAPI and Flask) that integrate with the Node.js backend through clean, production-grade APIs.

Built with React, TypeScript, Node.js, Express, PostgreSQL, and Python-based AI services, Centery is designed as a production-ready monorepo with workspace isolation, JWT authentication, scalable backend architecture, and real-world educational use cases.

ملفات مرفقة

بطاقة العمل

اسم المستقل
عدد الإعجابات
0
عدد المشاهدات
1
تاريخ الإضافة
المهارات