Overview

This document describes the software architecture of the AutoSuggestion Quiz system and explains how the requirements are mapped into the design.

The architecture is designed to:

• Support critical evaluation of AI-generated code suggestions
• Provide secure role-based access control
• Maintain low-latency code suggestions
• Log all meaningful interactions
• Store all persistent data in a shared database

---

Component Descriptions

Web Application (React/Vite)

Type: Client Interface

Purpose: Provide interactive UI for all user roles.

Responsibilities:

• Login and registration
• Student code editor and terminal
• Quiz interface (AutoSuggestion mode)
• Teacher dashboard (problem creation, grading)
• Admin panel

Interface:

• Communicates with Backend API using REST over HTTPS.

Backend API Server (FastAPI or Node.js)

Type: Application Logic Layer

Purpose: Central control of business logic, authentication, permissions, and coordination between services.

Responsibilities:

• Authenticate users (Edu ID or email/password)
• Enforce role-based access control (Student / Teacher / Admin)
• Manage classes, problems, quizzes, and access keys
• Handle submissions and grading
• Request AI-generated suggestions
• Send code to Code Runner for execution
• Log system events and interactions
• Provide analytics endpoints

Dependencies:

• Shared SQL Database
• AI Suggestion Engine
• Code Runner Service

Interface:

• REST API (JSON over HTTPS)
• Example endpoints:
  • POST /auth/login
  • POST /quiz
  • POST /run
  • POST /suggest
  • GET /analytics

AI Suggestion Engine

Type: Intelligent Service Component

Purpose: Generate context-aware next-line code suggestions.

Inputs:

• Problem description
• Student's current code
• Programming language
• Cursor position / context

Outputs:

• Correct next-line suggestion
• Distractors (0–2 plausible incorrect options)
• Optional explanation

Performance Considerations:

• Asynchronous API calls
• Timeout handling
• Optional caching of repeated prompts

Code Runner Service

Type: Execution Service

Purpose: Safely compile and execute student code in a sandboxed environment.

Supported Languages:

• Python
• Java
• C
• JavaScript

Responsibilities:

• Compile (if necessary)
• Execute code with test cases
• Return stdout, stderr, exit code, runtime
• Enforce memory and time limits

Security Considerations:

• Code runs in isolated environment
• Resource usage restrictions applied

Shared SQL Database

Type: Persistent Storage Layer

Purpose: Maintain all system data.

Stores:

• Users (students, teachers, admins)
• Classes and enrollments
• Problems and quizzes
• Submissions
• Grades and feedback
• Suggestion events
• Event logs

Design Goals:

• Single source of truth
• Indexed queries for performance
• Referential integrity via foreign keys