Know your weaknesses. Fix your habits. Let AI build your study plan.
A full-stack LLM-powered academic coaching web application that analyzes a student's grades and study habits, then generates a fully personalized performance report including AI-written recommendations, a 7-day revision schedule, motivation, and warnings when academic risk is detected.
This project combines two intelligence layers:
1. Rule-Based Classification Engine (Python) Grades are automatically classified into performance levels no LLM needed for this step:
- 🔴 Weak : grade < 10/20 → highest revision priority
- 🟡 Average : grade 10–13/20 → needs reinforcement
- 🟢 Good : grade ≥ 14/20 → consolidation focus
A 7-day revision plan is then generated programmatically, assigning days based on priority level.
2. LLM Text Generation (GPT-3.5 Turbo via OpenRouter) The classified data and generated plan are passed to GPT-3.5 Turbo, which writes the human-facing content:
- 📋 Personalized summary : addresses the student by name, references their actual subjects
- ✅ 4 concrete recommendations : 2 for weak subjects, 1 for study habits, 1 strategy tip
⚠️ Warning : only triggered if average < 9 or habits reveal serious risk patterns- 💪 Motivation : one punchy closing message to energize the student
- 👤 Personalized by name : the AI addresses the student directly
- 📊 Grade input : add any number of subjects with grades out of 20
- 🧠 Study habits quiz : 4 questions covering study hours, revision frequency, sleep, and techniques
- 🔢 Rule-based classifier : instant subject classification without LLM cost
- 🗓️ Auto-generated 7-day plan : weak subjects get more days, Sunday is always a general review
- 🤖 LLM-powered report : GPT-3.5 Turbo writes warm, specific, actionable feedback
⚠️ Smart warnings : only shown when genuinely needed, not by default
| Grade Analysis | Recommendations | Schedule & Motivation |
|---|---|---|
 |
 |
 |
student-performance-assistant/
├── main.py # FastAPI backend
│ ├── /analyze # Main endpoint classification + LLM report
│ ├── /health # Health check
│ └── /sample-request # Example request shape for frontend devs
├── frontend/
│ └── src/
│ └── pages/
│ ├── Home.jsx # Landing + name input
│ ├── Quiz.jsx # Grades + study habits (3-step flow)
│ └── Report.jsx # Full AI report display
├── screenshots/
└── .env # API keys (not included)
- User enters their name
- User adds subjects and grades (out of 20)
- User answers 4 study habit questions
- Rule-based engine classifies each subject and builds a 7-day revision plan
- Classified data + habits are sent to GPT-3.5 Turbo via OpenRouter
- LLM generates a warm, personalized report (summary, recommendations, warning, motivation)
- Full report displayed in 3 sections: Grade Analysis → Recommendations → Schedule + Motivation
- Python 3.10+
- Node.js 18+
- OpenRouter API key (free at openrouter.ai)
python -m venv venv
source venv/bin/activate # Windows: venv\Scripts\activate
pip install fastapi uvicorn python-dotenv requests pydantic
uvicorn main:app --reloadcd frontend
npm install
npm run devCreate a .env file in the root:
OPENROUTER_API_KEY=your_openrouter_key
| Method | Endpoint | Description |
|---|---|---|
| POST | /analyze |
Full analysis classification + LLM report |
| GET | /health |
Service health check |
| GET | /sample-request |
Example request body |
| Layer | Technology |
|---|---|
| Frontend | React + Vite |
| Backend | FastAPI (Python) |
| LLM | GPT-3.5 Turbo via OpenRouter API |
| Classification | Custom rule-based engine (Python) |
| Planning | Algorithmic 7-day scheduler (Python) |
| Environment | Python venv |
Sakni Tasnim
Telecommunications & Computer Engineering Student
🔗 GitHub • LinkedIn
Feel free to use, modify, and build on this project.