BobCourse-AI — University Course Evaluation System

A full-stack university course evaluation system built with React + TypeScript, Python FastAPI, and Java Spring Boot.

Architecture

React (Vite + TS)
↓ JWT auth
Python FastAPI (core)
↓ httpx (no student_id)
Java Spring Boot (analytics)
↑ pure computation, no DB
Python FastAPI ← PostgreSQL (exclusive)

Quick Start

Prerequisites

• Docker & Docker Compose
• Node.js 18+ (for frontend development)
• Python 3.11+ (for local backend development)
• Java 17+ (for analytics service development)

Run with Docker Compose

cp .env.example .env
# Edit .env — set a strong SECRET_KEY
docker compose up --build
# Initialize database schema and demo data
docker compose exec core alembic upgrade head
docker compose exec core python -m app.db.seed

Service	URL
Frontend	http://localhost:5173
Python API	http://localhost:8000
Java service	http://localhost:8080
PostgreSQL	localhost:5432
API docs	http://localhost:8000/docs

Demo Credentials

Role	Email	Password
Admin	admin@bobcourse.edu	Admin1234!
Instructor	instructor@bobcourse.edu	Instructor1234!
Student	student@bobcourse.edu	Student1234!

⚠ These are for demonstration only. Never use these passwords in production.

Local Development

Python Backend

cd backend/core
python -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install -r requirements.txt
# Run tests
python -m pytest --tb=short -v
# Start dev server (requires PostgreSQL running)
uvicorn app.main:app --reload

Seed the database

cd backend/core
python -m app.db.seed

Alembic migrations

cd backend/core
# Apply migrations
alembic upgrade head
# Generate new migration after model change
alembic revision --autogenerate -m "description"

Java Analytics Service

cd backend/analytics
./mvnw spring-boot:run # or: mvn spring-boot:run
./mvnw test# run tests

Maven is not required locally if using ./mvnw (Maven Wrapper).

Frontend

cd frontend
npm install
npm run dev # development server
npx tsc --noEmit # type check
npm run build # production build

Project Structure

BobCourse-AI/
├── backend/
│ ├── core/ # Python FastAPI service
│ │ ├── app/
│ │ │ ├── api/ # Route handlers
│ │ │ ├── ai/ # Responsible AI module
│ │ │ ├── core/ # Config, JWT, deps, security
│ │ │ ├── db/ # Base, session, seed
│ │ │ ├── models/ # SQLAlchemy ORM (15 tables)
│ │ │ ├── schemas/ # Pydantic schemas
│ │ │ └── services/ # Business logic
│ │ ├── alembic/ # DB migrations
│ │ └── tests/ # pytest test suite
│ └── analytics/ # Java Spring Boot service
│ └── src/
│ ├── main/java/ # DTOs, services, controllers
│ └── test/java/ # JUnit 5 tests
├── frontend/ # React + Vite + TypeScript
│ └── src/
│ ├── components/ # Shared components
│ ├── lib/ # apiClient, auth helpers
│ └── pages/ # Login, dashboards, forms
├── docs/ # All documentation
├── docker-compose.yml
├── .env.example
└── .github/workflows/ci.yml

Running Tests

Verified results:

• Python Core: 69 passed
• Java Analytics: 16 passed
• Total automated tests: 85 passed
• Frontend ESLint: 0 errors, 0 warnings
• Frontend production build: successful

# Python (no DB required)cd backend/core && python -m pytest --tb=short -v
# Java (requires Maven)cd backend/analytics && ./mvnw test# Frontend TypeScript checkcd frontend && npx tsc --noEmit
# Frontend production buildcd frontend && npm run build

Security

• All passwords hashed with bcrypt (cost factor 12)
• JWT HS256 tokens, 60-minute expiry
• Role-based access control on every protected endpoint
• student_id never returned to instructor-facing endpoints
• student_id never sent to Java analytics service
• Configurable minimum-response threshold before instructors can view analytics or export CSV reports
• CSV reports contain aggregated statistics only, with no individual submission indexes
• PII masking before any AI analysis

Documentation

Document	Description
Architecture	System design and component interactions
Database Schema	All 15 tables with relationships
API Documentation	All endpoints with request/response
Security	Security model and RBAC
Responsible AI	AI principles, PII protection, offline fallback
Bob Usage Report	How IBM Bob was used in this project
Human Judgment	Engineering decisions and rationale
SDLC Evidence	Phase-by-phase development evidence
Testing Strategy	Test approach and coverage
Technical Documentation	Setup, deployment, configuration