Tutor-AI

# 🎓 Tutor AI

An AI tutoring system with a FastAPI backend and a modern React (Vite + TypeScript + Chakra UI) frontend. It creates personalized study content, generates question sets, evaluates answers with feedback, and tracks progress. MongoDB backs the data layer, and optional Atlas Search + Vector Search enables semantic reuse and speed.

## ✨ Why Tutor‑AI (value in one minute)

Most AI study sites stop at “generate notes” or “ask a bot.” Tutor‑AI is built as a full learning workflow: it produces structured lessons, turns them into balanced assessments, evaluates your answers with constructive feedback, and tracks progress—all with privacy controls and aggressive cost/latency optimization.

What makes it more valuable than typical alternatives:
- 📚 **Learn better**: structured lessons (overview → key concepts → examples → applications → tips → summary).
- ✅ **Assess fairly**: Bloom/difficulty‑aware questions; MCQs auto‑normalized to 4 clear options with mapped answers.
- 💡 **Get real feedback**: concise explanations plus optional study suggestions, not just a score.
- 🔒 **Safe, fast, and cost‑smart**: privacy/PII safeguards and semantic reuse that cuts LLM calls; resilient fallbacks.

## Preview

Here’s a quick preview of the dashboard:

![Preview Video](video.gif)  

## 📚 Contents

- [For Users](#for-users)
- [Why Tutor‑AI](#why-tutorai-value-in-one-minute)
- [For Developers](#for-developers)
- [Troubleshooting](#troubleshooting)
- [License](#license)

---

## 👥 For Users

### 🌟 How it's different (at a glance)

- 📖 **Learn with structure, not noise**: generated content is readable, skimmable, and designed for revision.
- 🎯 **Questions fit your goal**: choose types, Bloom levels, and a difficulty mix; we enforce quality on MCQs.
- 🎓 **Feedback that teaches**: beyond correct/incorrect, you get short explanations and optional study tips.
- ⚡ **Quicker results over time**: similar topics reuse cached, verified material to save time and tokens.
- 🔐 **Saf