UniGuard AI is an academic early warning and advisor decision support system designed to detect student academic risk at an early stage using explainable predictive analytics.

Universities typically detect academic failure too late, after performance collapse has already started.

UniGuard AI provides:

• early academic risk detection
• student performance forecasting
• explainable risk reasoning
• automated advisor intervention plans
• group-level analytics dashboards

The system is designed as a local web application with a browser frontend and a Python backend.

Local web app model:

Browser UI
  ↓
FastAPI backend
  ↓
Risk engine
  ↓
SQLite database

Core modules:

• Risk scoring engine
• Academic forecasting
• Advisor dashboard
• Action plan generator
• AI recommendation chat

• 500 simulated students
• 15-week academic timeline
• Multi-subject evaluation
• Risk clustering (LOW / MEDIUM / HIGH)
• Explainable intervention planning
• Advisor analytics interface

Backend:

• Python 3.12+
• FastAPI
• SQLite
• OpenPyXL

Frontend:

• Vanilla JavaScript
• Local SPA architecture
• Canvas analytics visualization

git clone https://github.com/milord-x/UniGuard-AI.git
cd UniGuard-AI

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

export PYTHONPATH=apps/backend
uvicorn uniguard.main:app --reload

If you use fish, activate the environment with:

source .venv/bin/activate.fish

Then open:

http://127.0.0.1:8000