Quartz 4

Application Outline

3 min read

Application Outline: AI-Powered Learning Portal

This document outlines the vision, architecture, and features of the AI-Powered Learning Portal.

1. Application Overview

The project is an adaptive learning portal designed to deliver personalized educational experiences. It empowers tutors to create and manage curricula while leveraging an AI engine (Gemini) to tailor content delivery, evaluate student performance, and provide actionable insights. The entire system will be containerized with Docker for easy deployment and scalability.

2. Core Concepts

  • Personalized Learning: The platform moves beyond a “one-size-fits-all” model. The AI analyzes a student’s progress and learning style to curate and generate content that specifically addresses their needs and knowledge gaps.
  • Tutor-in-the-Loop: The AI serves as a powerful assistant to the human tutor. By automating evaluation and generating detailed progress reports, it allows tutors to focus their attention on providing high-level guidance and targeted support.

3. Key Features

3.1. Student Portal (Port 2001)

  • Authentication: Secure registration and login.
  • Dashboard: View enrolled courses, upcoming lessons, and overall progress.
  • Personalized Lessons: Access course materials that are dynamically curated and supplemented by the AI.
  • Interactive Quizzes: Take tests with various question types (multiple-choice, open-ended).
  • Instant Feedback: Receive immediate, AI-generated feedback on quiz submissions.
  • Progress Tracking: View a history of scores and completed lessons.

3.2. Tutor Portal (Port 2000)

  • Authentication: Secure registration and login.
  • Dashboard: Overview of all managed courses and student activity.
  • Curriculum Management (CRUD):
    • Create, update, and delete courses.
    • Add, update, and delete lessons within courses.
    • Upload learning materials (e.g., PDFs, and later audio/video) to be processed, chunked, and stored in the vector database.
  • Assessment Creation: Build quizzes and tests for lessons/courses.
  • Student Monitoring: View a list of enrolled students and their progress.
  • AI-Generated Reports: Access detailed reports on individual student performance, highlighting strengths, weaknesses, and areas needing attention.

3.3. AI Engine (Gemini) Features

  • Content Curation: On-demand, semantically search the vector database for the most relevant learning materials based on the curriculum and the student’s immediate needs.
  • Content Generation: Generate new, supplementary materials, such as customized examples, summaries of complex topics, or practice questions.
  • Student Evaluation: Automatically grade quizzes, including analyzing open-ended answers for conceptual understanding.
  • Report Generation: Analyze a student’s complete performance history (scores, time spent, topics failed) and synthesize a comprehensive, natural-language report for the tutor.

4. Technical Architecture

  • Deployment: Docker Compose
  • Backend: Python (FastAPI) or Node.js (Express)
  • Frontend: Two separate web applications (e.g., built with React, Vue, or Angular).
  • Relational Database: MariaDB (for structured data like users, courses, scores).
  • Vector Database: Qdrant (for semantic search of learning materials).
  • AI Integration: Backend service will communicate with the Gemini API.

5. Data Models (Entities)

  • User: Stores login credentials and role (student/tutor).
  • Profile (Student/Tutor): Stores user-specific details.
  • Course: Represents a course created by a tutor.
  • Lesson: A specific lesson within a course.
  • LearningMaterial: Represents a chunk of text content stored in Qdrant.
  • Quiz/Test: An assessment tied to a lesson or course.
  • StudentAttempt: Records a student’s submission for a quiz, including their answers, the AI’s feedback, and the final score.

6. High-Level API Structure

  • /auth: Endpoints for user registration and login.
  • /students: Endpoints for student-facing features (accessing courses, lessons, submitting quizzes).
  • /tutors: Endpoints for tutor-facing features (managing curriculum, viewing reports).
  • /ai: Internal endpoints for triggering AI tasks (content generation, evaluation).

This application provides the technical foundation for the EnrichedPath homeschool system. The content and curriculum structure for this application is defined in the Proposed Curriculum. For Phase 4 implementation details, see Phase 4: Homeschool Platform Implementation.

Graph View

Backlinks